Kroger Company

Text (OCR of the scan at left; may contain errors)

IN THE ATTER THE KROGER COMPANY l"INAL ORDER, OPINION, ETC., IN REGARD TO ALLEGED VIOLATION OF SEC. 5 OF THE FEDERAL TRADE COMMISSION ACT Docket 9102. Complaint, July 1977 Final Orde' Sept. 19B1 This order requires, among other things, a Cincinnati, Ohio, operator of a major retaij food chain to cease advertising survey-based food price comparisons which refer to any geographic area or competitor unless the survey that forms the basis for such claims is designed and executed in a competent and reliable manner. The company is required to ensure that employees responsible for pricing merchandise are not aware of the items selected for the survey prior to its completion, and that items to be surveyed be selected either randomly or using a representative sample of the products to be covered by the ad claims. Appearances For the Commission: Eddie Correia, Paul Eyre, William Jacobs Noble ,Jones and Wille L. Greene.

For the respondent: Stuart J. Land, Melvin Spaeth, Kenneth A. Letzler, Robert S. Thorpe James A. Beat, Jerold M. Block and Vicki J. Divoll, Arnold Porter Washington, D. COMPLAINT The Federal Trade Commission, having reason to believe that The Kroger Company, a corporation, respondent, has violated the provisions of Section 5 of the Federal Trade Commission Act, and that a proceeding by it in respect thereof would be in the public interest, issues this complaint:

PARAGRAPH 1. Respondent is an Ohio corporation with its principal office at 1014 Vine St., Cincinnati, Ohio. PAR. 2. Respondent, through its wholly-owned subsidiaries, is engaged in the operation of approximately 1 250 retail food stores in 20 states. In 1975, respondent ranked third in terms of sales in the United States among companies operating retail food stores. In the operation of these stores, respondent sells an extensive line of items including food, groceries, and other merchandisc. Respondent causes said items to be transported from its places of business in various States of the United States to its stores located in various other States ofthe United States. (2) PAR. 3. Respondent maintains, and at all times mentioned herein has maintained, a substantial course of trade in the production .., , Complaint 98 F.

processing, distribution, advertising, and sale of the aforesaid items in or affecting commerce, as "commerce" is defined in the Federal Trade Commission Act. Respondent has been, and is now, conclucting an advertising program known as the Kroger Price Patrol in or affecting commerce, as commerce is defined in the Federal Trade Commission Act.

PAR. 4. As part of its advertising program, respondent is, and has been, engaged in advertising results of price surveys generally represented to the consumer as the "Kroger Price Patrol." The Price Patrol compares, on a weekly basis, the prices of the products of respondent' s competitors with respondent' s stated prices for approximately one hundred fifty (150) items. The results of this comparison are advertised in the format of tables reflecting the number of respondent' s prices which are lower than higher than " or "the same as" the competitors' prices.

Typical of the substance of these tables is the following: (3) ,," e" - '"

ln' I I '''' \ .

'IHr. h.tuul'J1 AJ.

639 Complaint PAR. 5. Through the use of Price Patrol advertisements, respondent has made, and is now making, statements with regard to the Price Patrol Survey and the Survey results which accompany the tables described in Paragraph Four. Among said statements are the following:

Kroger is the Low Price Leader! The Price Patrol Proves It. The Price Patrol Proves You Save More at Kroger. (4) Documented Proof: Kroger Leads in Lower Prices! Price Patrol Proof, Kroger Leads with Low Prices. Shopping at Kroger wil enable you to spend less for your food than at. any other store.

PAR. 6. By the use of the tables described in Paragraph Four and the statements described in Paragraph Five, and others of similar meaning, respondent has represented, directly or by implication that:

A. The Price Patrol Survey is a methodologically sound survey. B. The results of the Price Patrol Survey prove that most items in respondent's stores are priced lower than in competitors' stores. C. The results of the Price Patrol Survey prove that shopping at Kroger, rather than at competitors' stores, will result in lower overall expenditures.

D. The results of the Price Patrol Survey prove that the amount of savings in overall expenditures by shopping at Kroger is directly related to Kroger s relative performance on the Survey. (5) PAR. 7. In fact:

A. The Price Patrol Survey is not a methodologically sound survey.

B. The results of the Price Patrol Survey do not prove that most items in respondent's stores are priced lower than in competitors stores.

C. The results of the Price Patrol Survey do not prove that shopping at Kroger rather than at competitors' stores will result in lower overall expenditures.

D. The results of the Price Patrol Survey do not prove that t.he amount of savings in overall expenditures by shopping at Kroger is directly related to Kroger s relative perfonnance on the Survey. 642 FEm;RAL TRADE COMMISSION DECISIONS Complaint 98 F.

Therefore, the tables, statements and representations set forth in Paragraphs Four, Five and Six were, and are, unfair and deceptive practices.

PAR. 8. The advertising statements described in Paragraph Five and the tables described in Paragraph Four, and others of similar meaning, represent, directly or by implication, that shopping at Kroger, rather than at competitors' stores, will result in lower overall expenditures.

PAR. 9. At the time respondent made the representations alleged in Paragraph Eight, it did not possess and rely on a reasonable basis the said advertisements arefor the representations. Therefore, deceptive or unfair. (6) PAR. 10. The advertising statements described in Paragraph Five, and the tables described in Paragraph Four, represent, directly or by implication, that at the time respondent made the representations alleged in Paragraph Eight, it possessed and relied on a reasonable basis for those representations, whereas in truth and in fact respondent did not possess and rely on a reasonable basis for said representations. Therefore, the said advertisements are deceptive or unfair.

PAR. 11. In the course and conduct of its Price Patrol advertising program, respondent has engaged in the following additional unfair and deceptive practices: respondent's Price Patrol advertisements fail to disclose that meat, produce, and house brands are not included in respondent' s Price Patrol Survey.

Thus, respondent has failed to disclose material facts which, if known to certain consumers, would be likely to affect their consideration to purchase respondent's items. Therefore, these practices were, and are, unfair and deceptive practices. PAIL 12. The use by respondent of the aforesaid deceptive and unfair practices has had, and now has, the tendency and capacity to mislead and deceive a substantial portion of the purchasing public into the purchase of substantial quantities of respondent's items. Therefore, these practices were and are to the prejudice and injury of the public and constitute unfair or deceptive acts or practices in or affecting commerce, in violation of Section 5 of the Federal Trade Commission Act.

THE KROGER CO.

639 Initial Decision INITIAL DECISION BY MONTGOMERY K. HYUN, ADMINISTRATIVE LAW JUDGE JUNE 11, 1979 PRELIMINARY STATEMENT On June 28, 1977, the Federal Trade Commission ("Commission issued a Complaint, charging The Kroger Company ("Kroger" or respondent") with unfair or deceptive acts or practices in violation of Section 5 of the Federal Trade Commission Act (15 V. C. 45) connection with certain advertising claims made in the retail sales of food products (the so-called Price Patrol" advertisements). By Answer dated August 24, 1977, as amended by Order Granting In Part Respondent' Motion For Leave To Amend And Denying Respondent' s Application For A Determination Permitting Interlocutory Appeal, dated November 15, 1977 , respondent denied that had violated the Federal Trade Commission Act as charged and interposed certain affirmative defenses. By Order Ruling On Complaint Counsel' Motion For (2jSl1mmary Decision, dated May 17 1978 (hereinafter sometimes referred to as "the Summary Decision Order ), and issued after oral argument, the administrative law judge determined that respondent had made certain representations in its Price Patrol advertisements as alleged in Paragraph Six A, B and C and Paragraph Eight of the Complaint. The parties were allowed extensive pretrial discovery. Prehearing conferences were held from time to time in order to simplify the issues, to resolve disputes related to discovery and generally to expedite the trial preparation of the parties. Two separate attempts at reaching a mutually acceptable consent order agreement undertaken in 1978 under my urging were unsuccessful. Evidentiary hearings were held from July 17 to July 25, 1978, in Washington, D. , for complaint counsel' s case-in-chief with respect to the remaining issues, including the issues of reasonable basis, disclosure of material facts and relief. Defense hearings were held from February 5 to February 15 1979 in Washington, D.C. The parties filed proposed findings, conclusions of fact and law, and proposed order, together with supporting briefs, on April 10, 1979 and answers on April 25, 1979. The proposed findings, conclusions and orders of the parties and their arguments in support thereof have been given careful consideration and to the extent not adopted by this Initial Decision, in the form proposed or in substance, are rejected as not supported by the evidence, irrelevant or as immaterial. Any motion appearing on the ) Initial Decision 98 F. record not heretofore or herein specifically ruled upon, either directly or by the necessary effect of the conclusions in this Initial Decision, are hereby denied.

Upon consideration of the entire record in this proceeding and having observed the demeanor of the witnesses, I make the following findings of fact and conclusions of law and order, based on the record considered as a whole: ' (3) FINDINGS OF FACT Introduction Identity of Respondent and the Nature of Its Business 1. The Kroger Company ("Kroger ) is an Ohio corporation with its principal offce at 1014 Vine St., Cincinnati, Ohio (Comp. TI 1, Ans. TI 1) 2. Kroger is engaged in the operation of approximately 1 170 retail food stores in 20 states. Kroger sells items commonly sold in such stores. Some of these items have been transported from Kroger s places of business in various States to some of Kroger stores located in various other States of the United States. Kroger has been ranked third in terms of sales in the United States among companies operating retail food stores (Ans. TI 2). 3. Kroger maintains, and at all times relevant to this proceeding has maintained, a substantial course of trade in the production processing, distribution. advertising, and sale of items commonly issold in retail food stores, in or affecting commerce, as "commerce" defined in the Federal Trade Commission Act (Comp. 11 3; Ans. TI 2; Ans. TI 3).

4. Kroger divides its retail food operations into Kroger Market- , 1"or the purpo!*s of this Initial Dedsion, the fonowing abbreviations were used. indings of Fact if) this Initial Dedsion. CPF Complaint couns'.' " Propose findings. CPHF Complaint counsel' s Proposed Reply Finding". RPF Respondent',; Proposed Findings Transcript of hearings, sometimes preceded by the name of the witness Complaint counsel's document.ary exhibit Respondent' s documentary exhibit.

Joint exhibit Comp - Complaint Ans Respondent's An wer to the Compaint Summary Decision - Order Ruling On Complaint. Colmse!' Motion far Summary Decision, date May 17, 1978 (hereinafter sometimes referred t the Summary Decision Order THE KROGER CO.

639 Initial Decision ing Areas ("KMAs ). There are 13 KMAs across the country: Central, Cincinnati, Columbus, Dallas, Delta, Erie, Gateway, Houston, Los Angeles, Louisvile, Michigan, Mid-Atlantic, and Southland (CX 1-1; JX 1: Pelln, p. 4; JX 1: Crague, pp. 3--). 5. (See In Camera Findings) 6. Kroger has conducted an advertising program, known as the Price Patrol, in or affecting commerce, as "commerce" is defined in the Federal Trade Commission Act (Ans. 3). B. Expert Witnesses 7. Dr. Frank Kohout testified as complaint counsel's expert on survey methodology. Dr. Kohout is an Associate (4JProfessor at the Department of Sociology, University of Iowa. Dr. Kohout holds a Ph.D. in sociology from Case Western Reserve University, Cleveland Ohio. Dr. Kohout has published numerous articles in learned journals and has presented papers before recognized professional societies regarding survey design and statistical methodology (CX 707). He has published a book on introductory statistics, entitled Statistics for Social Scientists designed to be used for either advanced undergraduates or for beginning graduate students (CX 707-3; Tr. 1002). In addition, he has written three chapters on statistics for a book called Statistical Package for the Social Science (Tr. 1001-02). Dr. Kohout also serves as a reviewer of articles published in recognized professional journals, including the American Sociological Review, Sociological Quarterly, and Sociological Focus and in a specialized journal called Symbolic Interaction. From time to time, Dr. Kohout has reviewed books for publishers (Tr. 997; CX 707). Dr. Kohout has performed consulting duties on research design, measurement, and statistics for a research firm, and has consulted for a number of manufacturers, state and federal agencies and for a federally funded research project (CX 707; Tr. 998). In his current position, Dr. Kohout teaches graduate-level courses in statistics and research methods (CX 707). He is a member of M. and Ph.D. review boards at the Department of Sociology, University of Iowa, and is responsible for advising advanced degree candidates with regard to survey design and statistical methodology (Tr. 997). Dr. Kohout was qualified, without objection, as an expert witness in the areas of research methodology, including survey design and applied statistics (Tr. J 002).

8. Dr. Robert Kleyle testified as complaint counsel's expert on statistical methods and their application. Dr. Kleyle is an Associate Professor of Mathematics and Statistics at Indiana Purdue University in Indianapolis. He has held that position for five years. Prior to Initial Decision 98 F. his appointment at Purdue, Dr. Kleyle was employed for 6 years as an Assistant Professor at the University of Massachusetts in Amherst, Massachusetts. Prior thereto, he was employed at Simmons College in Boston as a part-time instructor (Tr. 582). Dr. Kleyle holds his Ph.D. degree from Harvard University in statistics. The subject of his doctoral dissertation was the area and foundation of statistical inference (Tr. 583). Dr. Kleyle s current duties at Indiana Purdue University are primarily teaching. He is involved in (5) research in statistical methodology and, to a large extent, is involved in consulting, both within and outside of the University. He also teaches courses in mathematics, probability. and advanced statistics (Tr. 583-84). Dr. Kleyle has three publications in the general area of foundations of statistical inference. The first two puhlications appeared in the Annals of Mathematics Stat ,tics, and the third in the Annals of Statistics. He has published a paper on statistical inference for census data in Communications in Statistics another paper in the Journal of Quality Technology, dealing with a maintain- E.E. Transactionsability test plan, and still another paper in the LE. on Reliability (Institute of Electrical and Electronic Engineers), a reliability theory journal (Tr. 585). Dr. Kleyle was qualified as an expert witness in the application of statistical techniques to analysis of data, including the use of Chi-square analysis and the use of general statistical inferences in research methodology (Tr. 586-87). 9. Mr. Lloyd Oliver testified as respondent's expert on survey methodology and certain effects of price advertising. Mr. Oliver is Vice President of Glassman-Oliver Economic Consultants, Inc. Washington, D.C. (Tr. 2486). Mr. Oliver has a B.A. degree, with a major in economics (Tr. 2490-91). Mr. Oliver completed graduate courses for an M.A. in economics in 1972, but has not completed his thesis and has not received his master s degree (Tr. 2491, 2540). Mr. Oliver received a law degree from American University in 1975 (Tr. 2491). Mr. Oliver was employed, for more than five and one-half years, by the Federal Trade Commission as a staff economist in the Bureau of Economics. Mr. Oliver served as Deputy Assistant Director of that Bureau for approximately one year, and left the Commission in 1976 (Tr. 2493). While at the Commission, Mr. Oliver also worked as lead staff economist in matters involving survey methodology and design and matters involving the economic effect of ,trade restraints (Tr. 2527-33). Mr. Oliver has no publication in any ecognized professional economic journal requiring peer review (see \respondent's Pretrial Memorandum, dated December 22 , 1978). Mr. Hiver has not served in an editorial capacity, or as an official referee r reviewer of articles submitted for publication, for any professional THE KROGER CO. 647 639 Initial Decision economic journal (Tr. 2539). His statistical course work is limited to three statistics courses at the undergraduate level (6J(Tr. 2542). Mr. Oliver has not taught any courses in statistics or survey design (Tr. 2541), and has not designed any retail food price survey (Tr. 2547). Mr. Oliver has not conducted any studies of the relationship between substantiation requirements and the behavior of firms (Tr. 2552). Mr. Oliver was not involved in any matters, while employed at the Federal Trade Commission, in which retail food advertising was the primary issue (Tr. 2547-49). Mr. Oliver was qualified as an expert witness with regard to survey methodology and the effects of advertising in the retail food industry (Tr. 2555). 10. Dr. Lee Benham testified extensively as respondent' s expert on survey methodology and likely economic effect of the relief sought by complaint counsel. Dr. Benham is an Associate Professor of Economics, Washington University, St. Louis, Missouri and holds a Ph.D. in economics from Stanford University (Tr. 2797-98). He currently teaches courses in labor economics, basic price theory, medical economics, and industrial organization (Tr. 2798-99). Dr. Benham has published 12 to 15 articles in recognized professional journals in the area of economics (Tr. 2799), and has served as a reviewer of articles for professional journals (Tr. 2801). Dr. Benham has been employed as an economic consultant in a number of governmental and private industry matters (Tr. 2801) In particular he has served as a consultant and expert witness for the Federal Trade Commission in the eyeglass advertising proceeding which resulted in a trade regulation rule limiting the extent to which trade associations and state legislatures may restrict the flow of commer cial information in the retail eyeglass industry. The Rule also places limits on disclosure requirements for eyeglass advertisements (Tr. 2803-08). Dr. Benham s work has been relied upon by the FTC in its regulation on eyeglass advertising, 43 FR 23992, 23995 (1978), and in public statements (Tr. 2808-10). Dr. Benham s work has also been cited with approval by the Supreme Court. Roles v. 81ale Bar of Arizona 433 U.S. 350, 377 (1977); Friedman v. Rogers 47 U. 4151, 4157 (1979) (Blackmun, J. dissenting). Dr. Benham, however has never undertaken any study of any aspect of the retail food industry (Tr. 2815), nor of comparative price advertising (Tr. 2817- 18).

Certain Kroger Personnel 11. Thomas M. Vaughn is Kroger s Director of Meat Procurement. He was employed by Kroger on a part-time basis in the early 1950' s and continued on a part-time basis until (7Jcompletion of his Initial Decision formal education in the early 1960' 98 F time basis for various assignments, inwhenstoreshe was hired On a work. Sometime in the early 1960' full- , store management Cincinnati Division Offce as Meat Sales Promoter, he Was assigned to Kroger 12. Dr. Richard Bere is the Vice (fr. 1882 85). Marketing Area. Dr. Bere was President of the Columbus employed manager, and became assistant buyer in thebyproduceKroger indepartment.1957 as storeHe later became a field specialist in fruits and vegetables various stores, with the produce managers or running, workingthe depart-with ments. He Was appointed Produce Merchandiser sion, in 1962, where he remained until, Minnesota Divi' ferred to the Grand Rapids Division. From1965.GrandHeRapidsWas then trans- Was transferred to the Cleveland Division as Produce Merchandiser for 86 or 88 stores. While in the Cleveland Division, Dr. Bere Director of Merchandising and held that position for two years. Dr. Bere Was then transferred to Kroger, he became tant, reporting to the Vice President s General Offce as a consul- Kroger. He remained in that, Produce Merchandising for appointed Vice President of ProducepositionMer for a year until he fall of 1977, Dr. Bere Was transferred to the handising in 1972. In Wasthe Area as Vice President of the Columbus MarketingColumbus Area (Tr. 1961- 65). Marketing 13. Charles L.

Kroger Company, Thomas, Jr. is a Group Vice President of the responsible KMAs. Mr. Thomas has been employedfor the bysUpervisionKroger for of28 7years.of thePrior14 to his current position year and two mon ths, which he has occupied for approximately Southland Marketing Area, Mr. Thomas Was the Vice Presiden t of One , headquartered in for approximately two years. Prior to that the Nashvile Tennessee the Atlanta Division, he Was Vice President of Prior to that time, he,Wasand Viceheld Presidentthat positionof Groceryfor aboutMerchandisingfour years. and Procurement for Kroger. He also had Grocery Merchandiser for the Atlanta held the positions of Manager; Assistant Buyer; and Flo Division; Grocery Sales Division (Tr. 2093- Manager of the Cincinnati 2093-B). 14. Dean A. Hicks is the Vice tration for Kroger. He is responsiblePresidentfor the ofreviewMarketingof Adminisprocedures and for recommending and implementing changes for policies and thos policies and procedures that tend to cross the lines of authority of the four merchandising departments. Mr. Hicks is responsible for monitoring Kroger s compliance with the Federal Trade Commissionorder regarding advertised product availability and pricing. He is responsible for merchandising research and development , (BJpackag- THE KROGER CO. 649 639 Initial Decision ing research and development, and the design and implementation of the Universal Product Code (Tr. 2306-07). Mr. Hicks was directly involved in developing and writing the Everyday Low Price Policy ("EDLP") for Kroger (Tr. 2312). He also has the primary responsibility for the Burgoyne Checks (Tr. 2426).

15. Kenton A. Gast is a Director of Grocery Procurement at Kroger, and has held that position for 6 'h years. He has been employed by Kroger for just under 29 years. He started at Kroger in 1954 as a management trainee. He then became a store company manager for about a year and a half. His next positions were produce trainee, produce field representative, field representative, and buyer trainee. These positions covered a period of about three or four years. Then he was a produce buyer for two years, and served as Assistant Merchandiser in Grocery, Grocery Buyer-Merchandiser for two years. Mr. Gast was then transferred to Cincinnati as a field manager, and, about three years later, went to the General Office, Grocery Merchandising, where he worked for about five years (Tr. 2173-74).

16. Ted R. Hoover is the Vice President of Sales and Operations for Rabold Brokerage Company, Incorporated. Prior to his current employment, Mr. Hoover was employed by Kroger for 23 years, from 1953 to September 1977. At Kroger, Mr. Hoover s last position was Grocery Merchandiser, from 1970 until 1977. The Grocery Merchandising Department is responsible for buying grocery products designing sales programs for grocery items, and setting retail grocery prices (JX 2: Hoover, pp. 113-14). Prior to that, in 1970, Mr. Hoover was Grocery Merchandiser in Grand Rapids for Kroger for about a year and a half. Mr. Hoover has also held positions as Grocery Merchandiser in Minneapolis; Assistant Merchandiser in Little Rock, merchandising representative, store manager, head stock clerk, store manager trainee, and a clerk in the store (Tr. 1560- 62).

17. Alan Joseph Stec is Advertising and Sales Promotion Manager for Kroger in the Southland Marketing Area. He has been employed by Kroger for 13 years. Mr. Stec has held the positions of Advertising Manager for the Little Rock Division, Advertising Assistant Manager for the Atlanta Area, Advertising Assistant in Grand Rapids, Advertising Assistant in the Chicago Division, Relief , and trainee (Tr. Store Manager in the Chicago Division, co-manager 1692-93). In his present position, Mr. Stec is responsible for all the advertising in the Southland KMA, except to the extent that certain , Inc., an responsibilities have been delegated to Campbell-Mithun advertising agency (CXs 4- , 5--). He evaluates all advertising and Initial Decision 98 F. recommends to the management group techniques (Tr. 1692-94). (9) advertising procedures and 18. Robert Pellin is a Grocery Merchandiser for Kroger in the Central Marketing Area. His duties are to generate the sales and the necessary gross profits according to Kroger objectives, and to direct thes s Marketing Area position, Mr. Pellin Was the Assistantoperations.Grocery PriorMerchandiserto his currentfor the Central Marketing Area. He had also worked in the Marketing Area, in the Peoria Division Central before that at Kroger, at Market s St. Louis Basket, and Marketing Area (JX 1: Pellin Division, now called the G'!teway 19. Harry McDaniels is pp. 3-5). . the Advertising Manager for the Central Marketing Area. He has held that position for about eight years. He has been employed by Kroger for 27 years. As Advertising Manager Mr. McDaniels is responsible for all advertising, including newspaper advertising and radio and television commercials. He the activities of the advertising programs supervises CXs 4-5, 5-B). (JX 1: McDaniels, pp. 2- 20. Charles R. Cnigue is a Director of Merchandising for the Gateway Marketing Area. He has worked for Kroger for 25 years. Mr. Crague started out as a trainee manager, produce merchandising representative, co-manager, store manager, zone buyer, grocery buyer, assistant grocery , grocery sales diser. He has also worked as Managerpromoter,of Operationsand grocery merchan- Manager, Director of Sales for Grocery, and Director of, MerchandiseMerchandis- ing of the Gateway Marketing Area. As Director of Merchandising for the Gateway Marketing Area SUpervision of five departments, Mr. Crague is responsible for the ments are: Advertising, Grocery, Producewithin the region. These depart- (JX 1: Crague, pp. 3-5). , Meat, and Delicatessen 21. David Charles Ticeson and Administration Office inis Krogera Senior Analyst in the Marketing responsible for the scheduling, coordinations General Offices and Burgoyne Checks Was, and tabulation of the (JX 2: 22. Arthur L. Ticeson, p. 60). Ferguson, Esq. is in-house counsel for Kroger (Tr. 739). G. Leonard, Esq. is a member of Kroger (Tr. 739-40). F. Leland s Legal Department Davis is the Vice President of Advertising for Kroger General Headquarters Office (CX Manager of Consumer Research in Kroger 2-1). C. Ohlinger is the 1). James A. Leroy is the Vice President of the s Headquarters Offce (CX Area (CX 2-2). James M. Boswell is the Director ofDeltaMer Marketing for the Delta Marketing Area (CX handisingGrocery Merchandiser for the Delta Marketing2). EdwinAreaA. Schuler(CX 2- is the 2). (10) THE KROGER CO. 651 639 Initial Decision W. Robert Knight is the Advertising Manager for the Delta Marketing Area (CX 2-2). B. G. Beaty is the KMA Vice President for the Central Marketing Area (CX 2-2). A. Wayne Smith was the Director of Merchandising until February 16, 1976 (CX 2-2). Newton Briggs is the KMA Vice President for the Dallas Marketing Area (CX 2--; Tr. 745). Jack Davis was theKMA Vice President for the Southland Marketing Area until June 1, 1975 (CX 2-5). 23. L. Hollin was employed by Kroger s General Offce in the Advertising Department (Tr. 740). Tom Brand was a Kroger store manager in Dallas during July 1975 to June 1976 (Tr. 740). George Robinson is employed in Advertising at Kroger s General Office Headquarters in Cincinnati, Ohio (Tr. 741). Leland Carawan was an Assistant Advertising Manager at Kroger in Nashville (Tr. 742--3). 24. Bob Crow was first a Grocery Merchandiser, and became Director of Operations in Dallas (Tr. 743). Bennett Hudson was Regional Vice President for the Dallas Marketing Area (Tr. 743--4). Larry Turney was the Public Relations Manager for Kroger in the Dallas Marketing Area (Tr. 744). Earl Blackshire was the Assistant Grocery Merchandiser for Kroger in the Dallas Marketing Area (Tr. 744). Russ Thorburn was the Assistant Advertising Manager for Kroger in the Central Marketing Area (Tr. 744). Bob Bachleda is the Grocery Buyer for the Central Marketing Area (Tr. 745--6). Tom Robins, also known as Robbie" was the Assistant Grocery Merchandiser for the Central Marketing Area (Tr. 746). D. Certain Employees ofCampbellMithun, Inc. 25. Marvin Goldstnith is the Vice President of Campbell-Mithun Inc., and advertising agency (Tr. 740). John Kroenig is an account executive at Campbell-Mithun (Tr. 740). Gordie Krimstein is employed by Campbell-Mithun in the Creative Department (Tr. 741). Dave Carter is Kroger Account Executive (Tr. 741--2). Marty Chasen was a Kroger Account Executive (Tr. 742). Bill Stein was a Campbell- Mithun employee (Tr. 742).

E. Miscellaneous General Findings 26. CX 16, entitled "Managing A Kroger Price Patrol Program dated April 14, 1975, is a document written by Mr. Al Stec Advertising Manager of the Southland Marketing Area, in conjunc. tion with two advertising consultants in Kroger s headquarters (Tr. 1815). It was disseminated (llJto the KMAs as an information guide should the KMAs decide to enter a Price Pa.trol Program (Tr. 1814- 15; CXs 5- , 73). The document sets forth the details of how to put , pp.

652 J.'EDERAL TRADE COMMISSION DECISIONS Initial Decision together, run and manage a Price Patrol Campaign (Tr. 98 P was disseminated through Kroger 760). CX 16 Mithun (JX 1: PelJn, p. 11). s advertising agency, Campbell- 27. CX 15 is a document entitled "

p. 11). It was sent to The Price Patrol" (JX all operating divisions by thel: Pellin headquarters, explaining to them how the Price Patrol works and Kroger what the Price Patrol Program 10-11) CX 15 contained other such materialswas (Tr.and1674-75;informational 1: fromPellin Kroger regions as sales charts into the program itself. CX 15, and how they progressed as they got ing Department under the Was put together by the Merchandis- Supervision of Charles Crague Pellin, p. 11). The document Was written to be used as a communicating with (JX 1: merchandising philosophy of Kroger merchandisers in the means of field on the basic put together for the purpose of setting(JX down1: Crague,and discussingpp. 11-12).face-It Was face with individuals the merchandising philosophy of the company and how the Price Patrol fit in with Kroger tophilosophy (JX 1: s total merchandising testimony to CommissionCrague,Exhibitp. 21).1 relate(Referencesto CX 15. inSeeMr. Crague , Tr. 305-08. Some Salient Aspects Of Kroger Retail Food Stores And Their Operation Composition of Kroger Retail Food Stores 28. Kroger divides the items sold in its retail food stores into four general product categories: Grocery, Meat sen (JX 1: Pellin, p. 41; CXs 4- , Produce, and Delicatesitem includes every item in the ,stores-2). thatA "groceryis not meat" or "commodity delicatessen (JX 1: Pellin , produce, or everything sold in the Produce, p. Department41; Tr. 737). "Produce and vegetables (Tr. " includes 1965). , primarily fresh fruits have neither been frozen "Freshnor processedmeat" includesby curingmeator itemssmoking,whichor some further process, such as fresh fresh lamb, and fresh veal (Tr. 1886).beef, fresh pork, fresh poultry, Kroger as a brand which Can only be "Private label" is defined by example, a Kroger label on an item is asoldKrogerwithinprivatea company.label itemFor (Tr. 2175). The technical or trade USe of the interchangeably with word "grocerynot in accord fI2jwith"commodity"the to exclude meat and produce is common sense understanding of "cery," which includes food items generally sold in a grocery store including meat and produce (Tr. 1117- gro- 18. grocery " in Webster;, Third International DictionaryAlso seeof"grocer"the En!?lishandLangua!?e, Unabridged (1961)). THE KROGER CO. 653 639 Initial Dccision 29. The evidence shows that meat items comprise about 21 % of Kroger s retail food store sales, produce items 70/0 delicatessen items about 2%, grocery items 69%. Private label items comprise 28% of Kroger s grocery item sales (CXs 4-3, 8-2). 30. In the Southland marketing area, meat items comprised about 21 y,% of Kroger s retail food store sales during 1976, fresh produce items about 70/0 delicatessen items about 20/0 grocery items about 69%, and private label items about 28.3% (CXs 4- , 5-2). 31. The reports published by Supermarketing Magazine in Seps salestember 1977 on food expenditures generally parallel Kroger in that about 22% of all items in an average shopping basket for a family of four are meat items (28% of food items), and 11 % produce items (14% of food items) (CX 810; Tr. 1013-28). These percentages are generally confirmed by the results of a Bureau of Labor Statistics Study published May 15, 1975 (CX 801). Seventeen percent to eighteen percent of Kroger s dollar volume of grocery items are direct-stare-delivered (DSD) and are not distributed through Kroger s warehouses (JX 1: Pelln, p. 64). Meat and produce account for a significant portion of Kroger s dollar volume business (JX 1: Hoover p. 56). Meat and produce also comprise a significant portion of consumer expenditures for food (Tr. 2985). 32. A typical Kroger store will stock approximately 10 000 to 000 items (JX 1: Hoover, p. 18; JX 1: Pellin, p. 23). B. Pricinf? of Kroger s Food Products 33. The grocery merchandiser for each KMA has the responsibility for setting retail prices on all Kroger grocery items in the KMA (CXs 4-4, 5-4, 24- , 25-1; Tr. 761; JX 2: Hoover, pp. 134, 177; JX 2: Stec, p. 103; JX 1: Pellin, pp. 21, 41-43; Tr. 748, 1658). Food prices change frequently (CX 22-1), and each week approximately 500 to 700 grocery items may change in price in a Kroger store (Tr. 1659- 60). (13) 34. During the period the Price Patrol Program was operated in the various KMAs, Kroger also conducted a "roller" program (CXs , 16, 30-3, 41- , 42-1). A "roller" is an item sold at a price reduced from regular retail price, yet producing better-than-average markfor the item or subgroup it represents (Tr. 2093-K; CX 15-). Kroger instructed the KMA Grocery Merchandisers to price "rolled" items below competitors' prices, saying "It does no good to ' roll' an item down 41 and still be 21 higher than a major competitor" (CX 15-9). The "roller" price reflects a manufacturer s allowance or discount to Kroger (Tr. 1590, 2093-K), which was generally "passed through," in whole or in part, to the consumer (CX 41-1). The reduced retail or Initial Decision 98 F. roller" generally lasts from 3 to 12 weeks (JX 1: Pellin, p. 40; Tr. 1591, 2093- , 2093-L). At any given time, about 150 to 350 items will be "rolled" in a Kroger store (Tr. 735, 1595; JX 1: Pelln, p. 65). Fresh meat and produce generally have not been included in manufacturers' allowance programs (Tr. 674-75), and meat, produce and delicatessen items were not included in Kroger s roller program (JX 1: Crague, p. 25). Items which were selected for inclusion in the roller program were considered higher-volume items and items recognizable to the consumer (Tr. 1675; JX 1: Crague, p. 27). The use of "rollers" by Kroger replaced in large part the offering of specials or deepcut features (CX 15-; Tr. 2093-L, 2093-N to 2093-0). The grocery merchandiser for each KMA had the responsibility for deciding which items should be rolled, and for what period of time (JX 1: Pellin, pp. 40, 43). Roller items were advertised more frequently than items which were not rollers (CXs 40-1, 36-7, 15-7; Tr. 2148). Roller items were designed to attract consumers into the store (Tr. 2148).

35. Since the criteria for selecting rollers and Price Patrol items are generally the same (Tr. 1675-76), there would be a higher percentage of roller items on the Price Patrol list than the percentage of rolier items throughout the store (Tr. 2735). 36. Kroger, through Mr. Hicks, attempted to show that placing rollers on the Price Patrol Survey had no effect on the Price Patrol Survey results (RX 980; Tr. 2360). However, RX 980, reflecting information for the last four weeks of 1977 for Dallas, Texas (Tr. 2414- , 2418), cannot be generalized for all periods in all areas. Furthermore, using the same assumptions Mr. Hicks used in preparing RX 980, the inclusion of rollers in the Price Patrol Program in other areas supports an inference that inclusion of roller items in the Price Patrol Survey significantly affected the outcome of the Survey (Tr. 2415-18). (14) 37. Prices of fresh meat change rapidly at Kroger stores, but quality of meat does not change nearly as quickly as does price (Tr. 1947).

C. Adoption of an Everyday Low Price Policy and Its Operation 38. In 1971 and the early part of 1972, Kroger stores were relatively high priced on . . . regular shelf prices throughout the entire store" and relied on heavy feature advertising on the weekends to attract customers (Thomas, Tr. 2093-E; Hicks, Tr. 2315). During this time period, consumers considered Kroger to be a highpriced store (Thomas, Tr. 2093-D).

THt' !:KUGMt cu. b:)O 639 Initial Decision 39. In May of 1972 the decision was made that, in order to foster long-term growth, Kroger would become an "everyday low price ("EDLP") merchant (Thomas, Tr. 2093- , 2099; Hoover, Tr. 1566). Under the EDLP policy as initially envisioned, Kroger stores were to be as low as or lower in price than their competitors in each geographic market (Thomas, '1r. 2099; Hoover, Tr. 1564; Helmsing, Tr. 771- , 800-02; ex 36-2; ex 47-2). Kroger stores in each market would "target" their prices on the lowest priced competitor in the market offering comparable merchandising services (Hoover, Tr. 1638; lIelmsing, Tr. 800-01; CX 15-2).

40. The EDLP policy was meant to apply "across the board on all items that were carried in the (Kroger 1 store" (Thomas, Tr. 2093-E), including fresh meat and produce items (Thomas, Tr. 2104; Bere, Tr. 2002-05; Vaughn, Tr. 1929-34; CX 47-1). Respondent' s witnesses stated that as the EDLP policy was implemented in many areas Kroger stores, on a "total mix" of goods throughout the store, sought to be and were lower in price than their competitors (Thomas, Tr. 2099; Hoover, Tr. 1690; JX 2: Hoover, p. 156; JX 1: McDaniels, p. 25; CX 47-2). However, Kroger s internal price checks showed that Kroger more often than not had higher prices for meat and produce items (F. 145-167 infra).

41. In late 1972 Kroger s top management instituted a program to monitor compliance with the EDLP policy (Hicks, Tr. 2319). Mr. Dean Hicks was selected to administer this program at Kroger headquarters and is responsible directly to the top management of Kroger for analyzing the monitoring surveys he administers and for ensuring full compliance with Kroger s EDLP policy (Hicks, Tr. 2310- 2319 2335 2348). (15) 42. After consulting with Kroger employees in the field and at headquarters as to the best means for monitoring Kroger s competitive position (Hicks, Tr. 2320-21), Mr. Hicks established a procedure whereby KMAs submit " full book checks" to him for selected markets on a quarterly basis for review and analysis (Hicks, Tr. 2342). Mr. Hicks also developed the "Burgoyne check " or "shopping basket survey," and selected Burgoyne, Inc., an independent marketing research company, to conduct these surveys under his general supervision (Hicks, Tr. 2320-21).

43. Kroger s EDLP policy was in effect in the four Kroger Marketing Areas relevant to this proceeding throughout the time the Price Patrol ran (Oliver, Tr. 2678- , 2686-87; Benham, Tr. 2874-76). Of the 284 Kroger stores which participated in the Price Patrol Program in the four KMAs, 74% (210 stores) were subject to Burg-oyne price checks and 70% (199 stores) were subject to Initial Decision 98 F. quarterly full book checks (RX 978; Tr. 2349-52). The evidence shows, however, that while the Price Patrol was being conducted in over 100 cities (CPF 48), full book checks were conducted in only 18 of these cities, and Burgoyne checks in only 22 cities (including the same 18 where full book checks were made) (RXs 947-964; RPF 117). Thus, in about 78% of the cities in which a Price Patrol survey was conducted, neither a full book check nor a Burgoyne check was made (CPRF 6).

44. With respect to those markets where Price Patrol surveys were conducted but Burgoyne surveys and full book checks were not respondent' s expert in survey design and methodology speculated that, had such price checks been conducted in those markets, the results would probably have been consistent with Price Patrol survey results, because of Kroger s adherence to its EDLP policy and because Price Patrol results were consistent with results of Burgoyne surveys and full book checks in other markets. However, the same witness admitted that he could not be sure whether the prices might be in fact lower, the same or higher (Tr. 2681-82). 45. During the Price Patrol period 306 Burgoyne surveys were conducted in 22 cities in the four KMAs involved in this case (RXs 925-946). Kroger had a lower dollar cost total on 1 419 of the 1 550 individual Burgoyne checks-91.5%. For the 306 dates the Burgoyne checks were conducted, Kroger had a lower dollar total than all competitors, or a lower dollar total than all but one competitor, 284 times-92.8% (HXs 925-946; RPF 115-116). (16) 46. Viewed in light of the results of each check (instead of dollar totals), in 1973, for the cities in the four KMAs involved in this case and running the Price Patrol at that time, Kroger conducted a total of 14 Burgoyne surveys. Of the 14 surveys, Kroger lost, or, in other words, did not have the lowest prices of those chains checked, 7 times or 50% of the total surveys. In 1974, there were 32 Burgoyne surveys. Kroger lost 11 of the 32, or 34% of the total surveys. In 1975, there were a total of 86 surveys. Kroger lost 29 of the 86, or 33% of the total surveys. In 1976, there were 86 Burgoyne surveys conducted. Of the 86 surveys Kroger lost 29, or 33% of the total surveys. For 1977, of a total of 57 Burgoyne surveys conducted that year, Kroger lost 19, or 33% of the total surveys. In 1978, there were a total of 11 Burgoyne surveys conducted, and of that total Kroger lost 2, or 18% of the total surveys. Thus, Kroger lost 33% of the Burgoyne surveys conducted from 1973 to 1978 (RXs 925-946; CPRF 7).

47. During the Price Patrol period, quarterly full book checks were conducted in 18 cities where Price Patrol was also conducted in THE KROGER CO. 657 639 Initial Decision the four KMAs (RXs 947-964). A total of 189 quarterly full book checks were conducted in these 18 cities. Kroger had more items with lower prices than did the competitor on 608 of the 656 individual full book grocery checks-92.7% (RXs 947-964; RPF 117). 48. When viewed individually with respect to each full book check, in 1973, there were 8 full book checks conducted, and of that total Kroger lost 5, or 62.5% of the total checks. In 1974, there were 18 full book cbecks conducted, and of that total Kroger lost 5, or 33%. In 1975, there were 45 full book checks conducted, and of that total Kroger lost 6, or 13%. In 1976, there were 62 full book checks conducted, and of that total Kroger lost 11, or 17%. In 1977, there were 43 full book checks conducted, and of that total Kroger lost 8, or 18%. In 1978, there were 10 full book checks conducted, and of that total Kroger lost 1 , or 10%. Thus, from 1973 to 1978, Kroger averaged a percentage loss of 25% of all full book checks conducted (RXs 947-964; CPRF 8-9).

49. For the 189 dates the quarterly full book check was conducted during the Price Patrol period, Kroger had more lower prices of items in the grocery department than competitors, except one, 183 times-96.8% (RXs 947-964; RPF 118). However, the evidence also shows that during the Price Patrol period, 1973-1978, Kroger lost on an average of 25% of the time in which full book checks were made (CPRF 9). (17) 50. The full book price check results for meat and produce items are submitted separately from the grocery department checks (Hicks, Tr. 2345; Oliver, Tr. 2672). Counts of items priced higher lower, or the same for meat and/or produce items sometimes are not recorded on the recap sbeet. During the Price Patrol period, 179 full book checks whose recap sheets included meat and/or produce were conducted in the relevant area. Totals of numbers of items, including meat and/or produce items, on which Kroger s prices were lower than, higher than or the same as a competitor were recorded for 523 separate checks against competitors. Kroger had more items with lower prices than did the competitor on 487 of these 523 full book checks-about 93. 1% (RXs 947-964; RPF 119). However, the evidence also shows that of the 179 full book checks conducted between 1973 and 1978, which included meat and/or produce, Kroger lost on the surveys on an average of 32.8% (CPRF 10). 51. The results of most Price Patrol surveys generally showed that Kroger had more lower prices on surveyed items than any , Tr.surveyed competitor (RXs 990-991 , Nos. (4H5), (11); Hoover 1601-02; JX 1: Pellin, p. 82; JX 1: Stec, pp. 102-03). 52. Two surveys of food prices in Atlanta conducted by the FTC Initial Decision 98 F.T.C. staft in June 1973 showed Kroger to be low overall (RX 225 camera pp. 11-12). The Burgoyne surveys conducted in Atlanta during the Price Patrol program showed Kroger with lower dollar totals than Big Apple, Big Star, Colonial, Winn-Dixie, Food Giant and K-Mart (RX 925). The full book checks showed Kroger with more lower prices, both for the grocery department and for all items, than Big Apple, Colonial, Winn-Dixie, Food Giant, and Richway (RX 947; RPF 129). However, the evidence also shows that during the same time period in Atlanta, Kroger lost to A&P 33% of the time Burgoyne surveys were conducted and lost to Treasure Island 50% of the time (CPRF 12).

53. A survey of food prices in Indianapolis conducted by a "public interest" group in December 1975 showed Kroger to be low overall (Sample, Tr. 484-86).

54. The evidence shows that of the five full book checks conducted in Indianapolis during the Price Patrol period, Kroger had more lower item prices than Standard, Eisner, Thrift- Mart, and A&P (RX 958). Of the 10 Burgoyne surveys conducted in Indianapolis during the Price Patrol program, Kroger had lower dollar totals than Marsh, Eisner, Standard, and Thrift- Mart (RX 939; RPF 132). (18)However, of the same 5 full book checks, Kroger lost to Marsh 3 out of 5 times surveyed (RX 958). Of the 10 Burgoyne surveys conducted in Indianapolis during the Price Patrol program, Kroger lost 9 times, or 90% of the time for all surveys conducted (RX 939; CPRF 13).

55. The three principal elements of competition in the food retail business are price, product quality and service, and price may not necessarily be the determining factor. Respondent recognized the fact that consumers formed pragmatic judgments about quality or value on the basis of their shopping experience over time, taking into account not only price but product quality (Tr. 1926 2002 2211-13). 56. The EDLP policy was designed to insure that Kroger stores are not only competitive in terms of price on a "total mix" of goods throughout the store but also offered "good value" and sought to project Kroger s image as a "value leader." (Tr. 1690; 2099; ex 47- RX 901-2).

57. In contrast, the Price Patrol program focused on the prices selected items for advertising purposes. The Price Patrol excluded fresh meat and produce categories and did not relate to the total mix of items throughout the store. The Price Patrol advertisements in " but stress low prices and seek toevidence do not mention "value project Kroger as a "low price leader.

. . . . .

THE KROGER CO. 659 639 Initial Decision II. Kroger s Price Patrol Program And Claims Contained In Certain Price Patrol Advertisements Challenged In The Complaint The Purpose of the Price Patrol ProfVam 58. In late 1972, Kroger instituted an "everyday low price policy EDLP") (Tr. 2093- , 2099). Mr. Dean Hicks, who was directly involved in developing and writing the nEveryday Low Price" policy for Kroger (Tr. 2312), testified that Kroger s everyday low price policy "will be such that over time the consumer can buy her supermarketing needs at as good as or better total value at Kroger as she can at any other comparable retailer" (Tr. 2313-14). He defined total value" to include price and quality, considered over time (Tr. 2314-15).

59. Other Kroger officials confirmed that being "competitive" did not mean being the cheapest place to shop (Tr. 771- , 1564). For example, Mr. Ted R. Hoover, who was (19Jgrocery merchandiser for Kroger s Southland Marketing Area from 1970 until 1977 when he resigned, agreed that " the policy. was not necessarily to be lower than everybody" (Tr. 1638-39). "That doesn t necessarily mean that you would be lower than every competitor in the market" (Tr. 1638). The Kroger manual explaining the Price Patrol (CX 15) also confirms that Kroger s policy was not to be the cheapest place to shop: " Kroger prices must be aB low as the lowest merchant in town (offering the same services). The key is 'offering the same services.''' (CX 15- , emphasis added). 60. The Price Patrol surveys were not used nor relied upon by Kroger officials to determine Kroger s competitive price position (Tr. 1649 2439). Mr. Hoover testified that "Price Patrol. . . was not used . . . to determine somebody s pricing posture. " (Tr. 1649). Mr. Dean Hicks, whose responsibility included monitoring Kroger s "every day low price program" (Tr. 2306-07) testified: Q. I understand that, but my question is more simple. Did you use the Price Patrol results to help you to monitor your "Everyday Low Price" posture? No.

Q. In other words, you had checks done weekly in many KMAs, price checks done weekly in many KMAs, and you didn t use them to determine your competitive position; isn t that right? A. That is correct. (Tr. 2439) The record evidence clearly shows that Kroger officials, including those with responsibility for insuring Kroger s competitive posture did not use the Price Patrol to monitor Kroger s prices. 61. There is evidence showing that Kroger, concerned about its , p.

Initial Decision 98 F. reputation as a high-priced store, conceived the Price Patrol campaign in part as a merchandising and advertising program which will change that reputation and effectively communicate Kroger every day low price policy" to the consumer (Tr. 761; CX 15, CX 30- 2; JX 1: Crague, pp. 11-12; CX 75-17). However, the record also clearly shows that the Price Patrol was meant to do more. As the Price Patrol emerged, emphasis was increasingly placed on demonstrating or "proving" to the housewives that (20)Kroger had the lowest prices in town (Tr. 1695; F. 109-111 infra). Furthermore Kroger clearly recognized the role of the Price Patrol as a means of preventing further erosion of its sales in certain areas and to increase its market shares and profitability in other areas (Tr. 760; CXs 15, 70- , 75-17).

Price Patrol Survey Procedures 62. The Price Patrol program started in Atlanta, Georgia in 1972 (RX 900-3). The other Marketing Areas followed the pattern set by the Atlanta experience in running the program (Tr. 751 , 756-57, 794 1563, 1601). Although each KMA had responsibility for implementing and operating the Price Patrol program in its area, the Price Patrol procedures in any given KMA followed the manuals distributed by the General Office and further incorporated past experiences , p. 7; cf, CX 75 with CXs 15in other KMAs ('1r. 1773; JX 1: Crague and 16). For example, the program in Dallas was based on what had been developed in Atlanta, and the Price Patrol program in Indianapolis (Central Marketing Area) was based on what had been done in Atlanta and Dallas (Tr. 756-57; CX 36-1). 63. The control which Kroger s General Office exercised over the Price Patrol program in the KMAs was demonstrated by several events. The manuals describing the program (CXs 15 and 16) were distributed by the General Office. Each KMA utilizing the Price Patrol program followed the basic guidelines set forth in CXs 15 and , and conducted the program in substantially the same manner (CXs 15, 16 75; Tr. 756-57, 794-95; JX 1: McDaniels, pp. 6-7; Tr. 1871). The General Office, including Mr. Lyle Everingham Executive Vice President of Kroger (JX 1: Pellin, pp. 12-13), and Mr. Gene Hoffman, President of the Kroger Company (JX 1: Pellin 13), was given a presentation regarding the Dallas Price Patrol program before the program was instituted in Dallas (CX 39-1). The General Office required that specific changes be made in Price Patrol advertising copy (CX 22), and, after August 1977, each KMA using the Price Patrol was required to submit copies of all advertising to Kroger s General Office (CX 23). Campbell-Mithun, Kroger THg KROGER CO. 661 639 Initial Dccision advertising agency (JX 1: Pellin, p. 11), was largely responsible for the television commercials, and assisted with weekly television advertising copy and production in Dallas (Tr. 757). Campbell- Mithun also had input into the decision to commence the Price Patrol program in Dallas (Tr. 757). Campbell-Mithun informed the Kroger officials in Dallas about the results of the Atlanta program and provided examples of television advertisements used in Atlanta (Tr. 758). Finally, the (21)decision to terminate the program was a corporate decision (F. 107).

64. Items to be included on the Price Patrol check lists were selected by the grocery merchandiser in each KMA (Hoover, Tr. 1573; JX 1: Crague, p. 41; JX 1: Pellin, p. 52). The grocery merchandisers utilized three criteria to select Price Patrol items: (1) items had to be "high-volume," high-recognition items, meaning items which were readily recognizable to the consumer and likely to be purchased frequently by consumers (Thomas, Tr. 2100; Hoover, Tr. 1574, 1581; JX 2: Hoover, p. 114; JX 1: Pellin, pp. 54-56; JX 1: Crague, pp. 39, 42);

(2) items had to be taken from "commodity" areas throughout the store (Hoover, Tr. 1576; Thomas, Tr. 2100-1; Oliver, Tr. 2609-13, 2634; RXs 920-924; JX 2, Hoover, pp. 114-16; JX 1: Pellin, pp. 54-55; JX 1: Craguc, pp. 39, 42; Stec, Tr. 1699). "Commodity" in this context means "grocery" category (F. 28);

(3) items had to be commonly available on the shelves to be checked at competing stores (JX 2: Stec, p. 86; JX 1: Pelln, pp. 53-54; JX 2: Hoover, p. 128; Hoover, Tr. 1575; Thomas, Tr. 2101; CX 15-12; CX 16-).

65. Items in fresh meat and produce categories were not included in the Price Patrol because of Kroger s belief that the quality on these items varied among different stores so as to make these items not comparable (Thomas, Tr. 2101; Hoover, Tr. 1574-76; JX 1: Hoover, pp. 55-56; JX 1: Pelln, pp. 68-9; JX 2: Stec, pp. 85- , 88- 90).

66. Similarly, private label items generally were not included in the Price Patrol because of Kroger s belief that the quality of private label products varied among different stores (Hoover, Tr. 1575; JX 1: Hoover, p. 55; JX 2: Hoover, p. 128; JX 1: Pelln, pp. 53-54). However such private label items as milk and eggs were included on Price Patrol because the quality of these items was thought to be essentially the same from store to store (Hoover, Tr. 1575-76; JX 1: Pelln, p. 54; JX 2: Stec, pp. 86-87; Morgan, Tr. 714). 67. The selection criteria requiring that items be high volume Initial Decision 98 FTC. items and representative of the "commodity" areas of the store were designed to obtain a mix of commodity items (22)one would normally find in a customer s shopping basket (JX 1: Pellin, p. 56). In order to identify items which fit these two criteria, Kroger personnel used a quarterly Kroger computer printout report, which detailed the unit movement of commodity items through each Kroger warehouse (Hoover, Tr. 1574; Thomas, Tr. 2100; JX 1: Hoover, pp. 8-10; JX 1: Crague, pp. 42-43; JX 1: Pellin, pp. 61- , 67). These computer reports divide goods into commodity categories, listing the unit movement of all items which fall into each commodity grouping (Hoover, Tr. 1574).

68. Kroger personnel reviewed the computer printouts and selected the highest volume items, excluding private label items from the various commodity categories (Hoover, Tr. 1574, 1665-66; JX 1: Hoover, p. 10). Items which were not the "top mover " or highest volume item in a commodity category, but which were, in the grocery merchandiser s judgment, highly recognizable to consumers were also included in the Price Patrol (JX 2: Hoover, pp. 123, 126; JX I: Hoover, pp. 12- 20- 55; JX 1: Pellin, pp. 55-56; Hoover, Tr. 1666-67; Helmsing, Tr. 773).

69. Kroger personnel utilized separate documents on movement of commodity items shipped directly by the vendor to the store (direct store delivery, or "DSD " items) which were not included on the computer printout (JX 1: Pellin, p. 67). Between 15 and 20 of the items on each Price Patrol check list were DSD items (JX 1: Hoover pp. 9- , 19-20).

70. Use of the computer printouts and other documents, as well as the experience of the personnel selecting the Price Patrol items helped ensure that the Price Patrol lists were tailored to the commodity purchasing patterns of each marketing a,ea (JX 1: Hoover, p. 4; Hoover, Tr. 1576-77; JX 1: Crague, p. 43). The Vice President of each KMA reviewed the list of items selected by the grocery merchandiser to ensure that the criteria for selection had been followed (Thomas, Tr. 2101).

71. Each grocery merchandiser compiled a list of about 600 commodity items (Thomas, Tr. 2100-01; Hoover, Tr. 1579; JX 1: Pellin, p. 57). The Advertising Manager of the KMA then divided the 600 items into four smaller lists of 150 items each, one of which was used for Price Patrol each week. Items from the various commodity groups were apportioned evenly among the four smaller lists (Stec Tr. 1699; JX 2: Stec, pp. 84-85).

72. The Price Patrol lists were reviewed by the grocery merchandiser and his assistants approximately once (23)a quarter (JX 1: . .

Tt1 .rI\\.JV.L" H '-'V.

639 Initial Decision Hoover, pp. 23-24). Changes were made in Price Patrol lists approximately twice a year, most frequently because an item was dropped by Kroger or by a number of its competitors, so that comparison was impossible (Thomas, Tr. 2127; Hoover, Tr. 1578-79; Stec, Tr. 1699-1700; JX 1: Crague, pp. 43-44; JX 1: Hoover, p. 22; JX 1: Stec, pp. 73- , 119). Other reasons for changing items were the introduction of new items into the market, changes in popularity of items, and seasonality of some items (Thomas, Tr. 2127-28; JX 1: Hoover, p. 22; JX 1: Crague, pp. 43-44).

73. Respondent's employees Mr. Hoover and Mr. Hicks testified that the representativeness of the Price Patrol lists is significantly increased by the fact that there are direct pricing relationships known as "parity" relationships, among many similar retail food store items (Hoover, Tr. 1584- , 1641; Hicks, Tr. 2353-58). To the extent that this principle is put into practice by Kroger, a change in the price of a Price Patrol item may be accompanied by similar price changes in a number of related items which are not on the Price Patrol lists (JX 2: Hoover, p. 159). However, the parity pricing relationship discussed by Mr. Hoover and Mr. Hicks does not say anything about the prices of items at any competitor stores (Tr. 1640 1642, 1648-9) and does not even apply consistently at Kroger, as shown by several examples taken directly from Kroger s own representative Price Patrol lists. Mr. Hoover, testified that knowing the price of one size of an item at Kroger and at a competitor s store would give you "some idea" about Kroger s price relationship to its competitor on another size of the same item, but " it would depend on whether or not all competitors use parity pricing, which in fact they did not" (1'r. 1640). Thus, the "parity" concept did not allow Kroger to assume that, if one size product appears on the Price Patrol list, a competitive posture to a larger or smaller size product not on the Price Patrol list could be determined (Tr. 1646-7). For example, Mr. Hoover testified that RXs 905, 906, and 907 were typical Price Patrol lists (Tr. 1639). He further testified that on these typical Price Patrol lists (RXs 905, 906, 907), Kroger s price competitive position on certain like items of different sizes was different (Tr. 1644-52), Kroger was lower priced on a quart of Purex bleach and higher priced on a gallon of Purex bleach (Tr. 1644).

74. One example of "parity pricing" involves different flavors of the same name-brand product-e.g., different flavors of Betty Crocker cake mix, or different flavors of (24)Hi-C fruit drink-which are generally priced the same. Thus the price of one Betty Crocker cake mix or one navor of Hj-C fruit drink represents the price of all ,, pp.pp.

Initial Decision 98 F. Betty Crocker cake mixes or all Hi-C fruit drinks of the same size (Hoover, Tr. 1583; Hicks, Tr. 2355; Oliver, Tr. 2638-39; JX 2: Hoover pp. 116, 159).

75. Another "parity pricing" relationship is created by the pricing of a larger size of an item to be, on a per-ounce basis lower in price than the next smaller size of that item. Thus, a 20-ounce bottle of a particular brand of ketchup is likely to be priced, on a per-ounce basis, less than the 14-ounce size of the same brand (Hoover, Tr. 1582; JX 2: Hoover, p. 159; Hicks, Tr. 2355). 76. The items on the Price Patrol lists account for approximately 19% of total grocery department sales and 11% of total retail store sales (RX 979; Hicks, Tr. 2352-54; Oliver, Tr. 2637). The total sales of Price Patrol items and items actually priced in parity with them are estimated by Kroger to account for approximately 42% of total grocery department sales and 25% of total retail store sales (RX 979; Hicks, Tr. 2352-54; Oliver, Tr. 2637).

77. The stores selected for inclusion on the Price Patrol were Kroger s major competitors those which were familiar to consumers in that geographic market and which operated comparable outlets with similar merchandising services (JX 1: Pellin, p. 69; CX 15- , CX 15-11; CX 16-5).

78. The Price Patrol price checkers were not full-time Kroger employees, but worked part-time and only on Price Patrol. Civic groups, television stations, store managers, and others suggested candidates who were then given screen tests at a television station. Those applicants who had good camera appearance were selected as price checkers (Stec, Tr. 1703; JX 1: McDaniels, p. 6; JX 2: Stec 77-78).

79. The Price Patrol price checkers were given thorough instructions on how to conduct the Price Patrol checks (JX 1: Pellin, pp. 96- 97; Stec, Tr. 1704; Helmsing, Tr. 763). As part of their training, they were " walked through" a Kroger store by the grocery merchandiser sometimes accompanied by the advertising manager, at which time the Price Patrol check list was reviewed item by item (Stec, Tr. 1704; Helmsing, Tr. 763; JX 1: Pellin, pp. 96-97; JX 1: McDaniels, p. 10; CX 16-, CX 16-6). (25) 80. The Price Patrol price checkers were free to check their assig.ned stores any day of the week they chose (JX 2: Stec, p. 79), although it was recommended that checks be made on certain days of the week when stores were more fully stocked (JX 1: McDaniels , 34-35; Helmsing, Tr. 765-66). They were not told which Kroger store or competitors' stores to check, and were free to check different stores each week (JX 2: Stec, p. 79; JX 1: Pellin, p. 69; JX 1: ,, p.p.

THE KKULir.t L-V.

639 Initial Decision McDaniels, p. 13). Prices in Kroger stores were checked in the same manner as in competitors stores, by recording retail prices as they appeared on items on the shelf (JX 1: Pellin, p. 70; JX 1: Crague 56), However, Kroger s own prices were occasionally obtained from company price books rather than from the actual shelf prices (CXs 25-2, 37- , 37- , 53).

81. The KMA Advertising Department designated which of the three or four lists was to be checked each week, rotating the lists from week to week (Stec, Tr. 1701-02; RX 903; JX 2: Stec, pp. 81-83; JX 1: McDaniels, pp. 12-13; JX 1: Crague, pp. 54-55). The order of rotation was determined at the beginning of each quarter for the coming 12 or 14 week period (Stec, Tr. 1700; RX 903; JX 1: Stec 120; CX 16-, 15-17). The Advertising Department sent the lists by mail directly to the price checkers (Stec, Tr. 1700; JX 1: McDaniels pp. 12 13; JX 2: Stec, p. 82; JX 1: Crague, p. 54). 82. Neither the grocery merchandising department, the store managers, nor anyone else other than the KMA Advertising Department and Price Patrol price checkers knew prior to each week' s Price Patrol check which Price Patrol lists would be checked that week (Hoover, Tr. 1581; JX 2: Hoover, pp. 119 20; Stec, Tr. 1702; JX 1: Stec, pp. 120 21; JX 2: Stec, pp. 81-83; JX 1: Crague, p. 54; JX 1: Hoover, pp. 64-65, 6&-9).

83. After completion of each week's price check, the check lists were returned to the Advertising Department (Stec, Tr. 1705). Trained Kroger advertising personnel then "recapped" the check lists, that is, calculated and recorded the number of items on which Kroger was priced lower than, priced the same as, and priced higher than, each competitor (RX 902; JX 1: Stec, pp. 90-91). 84. Emphasis was placed on ensuring that the Price Patrol checks were accurate (Stec, Tr. 1707 , 1824; RX 902; JX 1: McDaniels, p. 11; CX 15-12; CX 15-5, 15-7; CX 95-2). (26JThe KMA assistant advertising managers, as well as the individuals recapping the price check lists, reviewed the lists for obvious errors (Stec, Tr. 1705 08; JX 1: Stec, pp. 84- , 89-91; JX 2: Stec, pp. 105 06). 85. When it was suspected that an error had been made, the pricing analysts in the grocery merchandising department were consulted. If it was confirmed that an error had been made, the price checker responsible was immediately so notified (RX 902; Stec, Tr. 1705; JX 2: Stec, p. 106; JX 1: Pellin, p. 95; JX 1: McDaniels, p. 11; CX 62). If a price checker made repeated errors, a grocery merchandising representative would take her to a Kroger store to instruct her again on the proper price checking procedures (RX 902). 86. Items on which errors had been made to Kroger s advantage "7-444 0 - 82 - 4 QL 3 , pp.

Initial Decision 98 F. were removed from that week's Price Patrol recap, but errors in favor of a competitor were left in the report (JX 1: Pel1in, p. 91). 87. The KMA assistant advertising managers also checked the recap sheets for accuracy in the computation of the number of items on which Kroger was priced lower than, higher than, and the same as each competitor (Stec, Tr. 1707). After the recapping process was completed, the results (in the form of a box score) were inserted in the television scripts and newspaper advertisements which were to publicize that week's Price Patrol checks (JX 2: Stec, pp. 80, 90-91; JX 1: Pellin, p. 91).

88. The recap sheets, along with exact duplicates of the check lists detailing the item-by-item price comparisons, were posted in each Kroger store (Thomas, Tr. 2106-07; Stec, Tr. 1709, 1752-55; JX 1: Pellin, p. 91; JX 2: Stec, p. 80; CX 16-10, 16-18; CX 74-10). RXs 905, 906 and 907 are representative of the recap sheets and Price Patrol check lists which were posted in Kroger stores (Stec, Tr. 1752- 55).

89. Customers were invited in Price Patrol advertising to examine the complete checklists (see RFP 201). A significant number of Kroger customers did so (Stec, Tr. 1710, 1814; CX 708-3). Representatives of Kroger s competitors also reviewed the recap sheets and price check lists posted in Kroger stores (Stec, Tr. 1710- , 1753). 90. When Kroger "lost" a weekly Price Patrol check, that is when a competitor was loy.'r priced than Kroger on a greater number of items, Kroger truthfully so stated in its advertising (see RPF 204). When that occurred, the (27Jgrocery merchandising department conducted a special full book or other price check of that competitor to ascertain what price adjustments, if any, Kroger needed to make in order to assure its prices were as low as its competition (Hoover, Tr. 1602--3, 1605, 1607-11; JX 1: Hoover , 52; JX 2: Hoover, pp. 178-80).

91. The 150-item list checked each week by the Price Patrol checkers was taken from a master list of approximately 400-B00 items (Tr. 764; CXs 15 , 81). The 400-600 items are selected by the Grocery Merchandiser in each markeUng region, and the master list was divided into four 150-item lists (CXs 4 , 5 , 21-8; Tr. 747 48, 761 , 764, 765; JX 1: Hoover, pp. 3--; JX 1: Crague, pp. 41--3). When the list is changed, the Grocery Merchandiser is responsible for deleting or adding items (Tr. 765; JX 1: Crague, p. 44; JX 1: Hoover, p. 23). The master list is not made available to the public or to competitors.

92. Generally, the advertising department in each KMA rotated the list to be used each week (CX 5 3; JX 1: Stec, p. 114; Tr. 1699). THE KROGER CO. 667 639 Initial Decision Initially, in Indianapolis the Grocery Merchandiser was responsible for selecting the master list and for breaking the list down into four weekly samples, and for selecting which list to use each week (JX 1: Pellin, pp. 52, 58, 102). In the Delta KMA, the advertising department, together with the grocery merchandising department, selected the rotation of the lists (CX 75-13).

93. In sum, the Price Patrol lists selected by the Grocery Merchandiser had the following general characteristics: a. No fresh meat items were included (Tr. 1335, 1574, 1661, 1815 2101 2144-45 2445 2708 2855; CXs 806, 807, 808). b. No produce items were included (Tr. 1335, 1574, 1661, 1815 2031 2101 2144-45 2445 2708 2855; CXs 806, 807, 808). e. No delicatessen items were included (Tr. 1661 2144-45, 2745- 46; CXs 4-1, 8-1).

d. The items selected on the lists were high-movement items (in terms of dollar volume through Kroger s warehouse), and items recognizable to the consumer (Tr. 1664, 1667; JX 1: Hoover, p. 8; JX 1: Crague, pp. 39, 42; CX 5-4). (28) e. The items selected were national brands, except for, occasionally, milk, ice cream, fresh bakery items, and other commodity private label items (CXs 1&-10, 75-13; Tr. 773-74, 1575, 2144-5; JX 1: Hoover, p. 13).

f. The items selected were not chosen on the basis of random selection (Tr. 1665-66, 2715).

g. The items selected were not based on the distribution of actual consumer food purchasing patterns in the city or region (Tr. 2731). h. The items were not selected in relation to Kroger s food store sales volume (Tr. 1667 2731).

i. Very few of the items were direct-store-delivered items (Tr. 1661-62 2144-5).

94. The grocery merchandisers were aware of which items were rollers" and which items were included on the Price Patrol lists (Tr. 1687, 2145-46). The criteria for selecting rollers were generally the same as the criteria for including items on the Price Patrol Survey (Tr. 167&-76). In fact, the Price Patrol guideline manual (CX 15) strongly suggests that "proper roller items" be included in the Price Patrol lists:

If the correct pricing posture is being maintained and jf the proper "roller" items are being used, Kroger will be lower on more items and higher on fewer items than any competitor checked. If this is not the case the television is wasted and cannot be used. (CX 15-13).

, pp.

Initial Decision 98 F. 95. Mr. Pellin, the Grocery Merchandiser from the Central Marketing Area, testified that he regularly changed items on the Price Patrol lists to include 20 to 40 items which were to be rolled down in price for that week (JX 1: Pellin, pp. 58-0, 72). The only factor Mr. Pelln considered in deciding to change the list was whether or not the item was rolled (JX 1: Pellin, p. 61). 96. Taking the maximum number of rollers in a Kroger store at one time (350) and dividing by the lowest estimate of the total number of items in a typical Kroger store (29)(10 000) (F. 32), it can be concluded that only 3.5% of Kroger items are rolled at any onc time. Yet, Mr. Pellin states that between 13% (20 divided by 150) and 26% (40 divided by 150) of the items on the Price Patrol were rollers (JX 1: Pelln, p. 60). Mr. Hoover stated that between 15% and 20% of the items on the Price Patrol in the Southland Marketing Area were rollers (Tr. 1597). Kroger s own document, RX 980, shows that in Dallas, the percentage of rollers on the Price Patrol ranged from 7% to 13% for the last four weeks of 1977.

97. Thus, rollers are represented on the Price Patrol in a percentage in excess of the percentage of roller items in the store. Mr. Oliver, Kroger s expert witness, agreed that one would expect the rollers to be over-represented on the Price Patrol (Tr. 2735). Mr. Oliver also conceded that this over-representation provides a bias in favor of Kroger (Tr. 2735-36).

98. Kroger excludes at least 50% of the items in the store from Price Patrol lists: meat (21 %), produce (7%), delicatessen (2%), most grocery private label items, 28% of grocery items or 19% (28% X 69%), most low-volume items in the store, and most direct-storedelivered items (17'7 18'7) (CX 5-2; F. 93). Therefore, Kroger could have higher prices than its competitors on meat, produce, delicatessen, most private label items, most direct-stare-delivered items, and most low-volume items in the store, and still win on the Price Patrol (Tr. 2144-45).

99. As initially conceived and actually executed, the Price Patrol program used television as its primary advertising medium (CX 30- , CX 30-5; CX 15-11; CX 47-1; Stec, Tr. 1712). During the period of Price Patrol, 60% of the advertising budget of the Southland KMA was used for television advertising and 20% for newspaper advertisements (Stec, Tr. 1712).

100. The media usage pattern utilized in Southland was generally followed in other KMAs which adopted the Price Patrol program (Stec, Tr. 1871-72; Helmsing, Tr. 756-58, 794; JX 1: McDaniels 20-21; ex 47-1).

101. The advertisements contained in RX 537-1 through 14 , pp.

THE KROGER CO. 669 639 Initial Decision (television commercials), RX 540-1 through 2 and RX 566-1 through 2 (newspaper advertisements) are typical of the Price Patrol advertisements.

102. The claims contained in the Price Patrol advertisements regarding the number of items on which Kroger was prjced lower than, higher than, or the same as each competitor each week were based solely on the Price Patrol price checks (JX 1: Hoover, p. 39; Stec, Tr. 1822-23). (30) 103. Kroger employees testified that the claims contained in the Price Patrol advertisements that Kroger was the "low price leader were based not only on the Price Patrol price checks, but also generally on Kroger s EDLP pricing policy and a number of other price checks including the full hook checks and the Burgoyne checks which in Kroger s view verified its competitive position as the lowest priced retailer in the market (Thomas, Tr. 2158; JX 1: McDaniels 22-23; Hclmsing, Tr. 774- , 77- , 796-99; Stec, Tr. 1695-96; JX 1: Pellin, pp. 17-18; Thomas, Tr. 2099-2100; Hoover, Tr. 1612 , 1690). However, they also admitted that Kroger did not use any specific full book check or Burgoyne check for the purpose of verifying the accuracy of Price Patrol advertising program (F. 205-206, 222-223 infra).

The Scope of the Price Patrol Advertising Program. 104. Kroger s Price Patrol program began in Atlanta, Georgia, in December of 1972 (UX 900-3). Since that time, Kroger introduced the program in six other KMAs, including over 100 cities in 14 states: Alabama, Arkansas, Georgia, Indiana, Kansas, Kentucky, Louisiana Mississippi, Missouri, Ohio, Oklahoma, Pennsylvania, Tennessee and Texas (CX 1-3 through 1-5; RX 900).

105. The following tables show where and when the Price Patrol program was conducted in the Central, Dallas, Southland, and Delta KMAs (CXs 1, 3; RX 900):

Initial Decision 98 FTC. Price Patrol Beginning Dates and Ending Dates Central KMA Be€"in End Indianapolis 9/8/75 9/11/76 Bloomington 10/20/75 9/11/76 Kokomo 10/27/75 9/11/76 Ft. Wayne 10/5/75 3/4/78 Dallas KMA Begin End Dallas Area Dallas 3/18/75 3/11/78 Irving 3/18/75 3/11/78 Duncanville 3/18/75 3/11/78 Richardson 3/18/75 3/11/78 Farmers Branch 3/18/75 3/11/78 Grand Prairie 3/18/75 3/11/78 (31) PIano 3/18/75 3/11/78 Garland 3/18/75 3/11/78 De Soto 3/18/75 3/11/78 Mesquite 3/18/75 3/11/78 Lewisville 4/22/75 3/11/78 N acogdochcs 3/18/75 3/11/78 Waxahachie 3/18/75 3/11 /78 Marshall 3/18/75 3/11/78 Paris 3/18/75 3/11/78 Greenville 3/18/75 3/11/78 Shreveport, Louisiana 3/18/75 3/11/78 Bossier City, Louisiana 3/18/75 3/11/78 Sherman/Denison 3/18/75 3/11/78 Gainesville 3/18/75 4/10/76 Nachitoches Louisiana 3/22/75 9/22/76 Ft. Worth 3/29/75 3/11/78 Arlington 3/29/75 3/11/78 Terrell 11/19/77 3/11/78 Denton 5/15/75 3/11/78 Tyler 5/8/76 3/11/78 Weatherford 6/18/77 3/11/78 Longview 3/22/75 3/11/78 Sulfur Springs 3/22/75 3/11/78 Hurst 8/24/76 3/11/78 Duncan, Oklahoma 6/26/77 2/25/78 THE KROGER CO. 671 639 Initial Decision Dall s KMA Begin End Ada, Oklahoma 6/26/77 2/25/78 Lawton, Oklahoma 6/26/77 2/25/78 Southland KMA Bef!in End Atlanta 12/7 /72 3/15/78 Athens 10/8/73 3/15/78 Augusta 6/17/73 3/15/78 Albany 7/22/73 3/15/78 Carrollon 11/30/76 3/15/78 Chatt:nooga 4/1/73 5/6/78 Macon 12/2/73 3/15/78 Nashvile 6/30/74 5/6/78 Knoxville 6/4/73 3/15/78 Huntsville 6/30/74 5/6/78 Decatur 6/30/74 5/6/78 Florence/Sheffield 7/28/74 5/6/78 Bowling Green 7/28/74 5/6/78 Gallatin 9/23/75 5/6/78 Douglasville 9/22/75 3/15/78 (32) elta KMA Humboldt, Tennessee Warren, Arkansas Paragould, Arkansas Parris, Tennessee Poplar Bluff, Missouri Brownsville, Tennessee Wynne, Arkansas Blytheville, Arkansas Hamburg, Arkansas Clarksdale, Mississippi Tunica, Mississippi Tupelo, Mississippi Marks, Mississippi Cleveland, Mississippi Starkville, Mississippi Camden, Arkansas Aberdeen, Mississippi Oxford, Mississippi Prinkley, Arkansas Greenwood, Mississippi Kosciusko, Mississippi Helena, Arkansas West Point, Mississippi Greenville, Mississippi England, Arkansas Vicksburg, Mississippi Dewitt, Arkansas Malvern, Arkansas Lonoke, Arkansas Hope, Arkansas Paris, Arkansas Stuttgart, Arkansas Morrilton, Arkansas Jonesboro, Arkansas Magnolia, Arkansas Newport, Arkansas Jackson, Tennessee Russellvile, Arkansas Initial Decision 98 F. Union City, Tennessee Batesville, Arkansas Paducah, Kentucky Conway, Arkansas Mayfield, Kentucky Pine Bluff, Arkansas Murray, Kentucky Hot Springs, Arkansas Dyersburg, Tennessee Ft. Smith, Arkansas Little Rock, Arkansas Memphis, Tennessee 106. For all cities in the Delta KMA wher 2 the Price Patrol was conducted, the Price Patrol began sometime during August 1975 and September 1975. The Price Patrol ended in all such cities by December 1 , 1976, with the Price Patrol ending in most such cities by September 1, 1976 (RX 900).

107. Most of the cities conducting the Price Patrol program did not terminate the program until after the Complaint issued in Docket No. 9102. The Price Patrol program was terminatcd in all cities by May 1978. The decision to terminate the program was a corporate decision, and was based primarily on the fact that this litigation (Docket No. 9102) was pending (Tr. 2130, 2161). Certain Claims Made in the Price Patrol Advertisements 108. The Kroger Price Patrol program was advertised in newspapers, circulars, and on television (JX 1: McDaniels, pp. 16-17; Tr. 1712-13). In the Southland Marketing Area, 60% of the advertising budget was spent on television (33Jadvertising for the Price Patrol (Tr. 1712), and 15% of the advertising budget on newspaper advertising, which included the Price Patrol (Tr. 1712). This relationship generally held true for other KMAs which conducted the Price Patrol (Tr. 1871). In the Southland Marketing Area, the production costs for advertising represent 13% of the advertising budget (plus or minus 2%) (Tr. 1878). As the Price Patrol program matured and sales increased, the media cost, as a percentage of sales, was projected to decrease (CXs 15- , 13). Kroger advertised the results of the " Kroger Price Patrol" on a weekly basis, comparing some Kroger prices with prices of the leading competitors in a particular geographic area. The results of these comparisons were advertised in the form of tables (Ans. n 4). These tables displayed results showing the number of items for which Kroger s prices were higher than lower than, or the same as, each competitor s prices but did not display the actual prices (Tr. 2765; JX 1: Pellin, p. 76; CXs 125, 131; RXs 905, 906, 907). Certain advertising claims accompanied the charts which purported to generalize from the survey results (Ans. n5).

Tile KROGER CO. 673 639 Initial Decision 109. Kroger has represented in certain advertisements, directly or by implication, that the results of the Price Patrol Survey prove that most items in respondent's stores are priced lower than in competitors' stores (Complaint Counsel's Motion For Summary Decision, dated February 24, 1978; Memorandum Of Respondent The Kroger Company, In Opposition To Complaint Counsel's Motion For Summary Decision, dated April 14, 1978; Reply To Memorandum Of Respondent, dated April 24, 1978; Reply Memorandum Of Respondent, dated May 6, 1978; Order Ruling On Complaint Counsel' s Motion For Summary Decision, dated May 17, 1978 (pp. 4- 6); Motion Of Respondent, The Kroger Company, For Reconsideration Of The Order Of May 17, 1978, Or, Alternatively, For An Interlocutory Appeal From That Order, dated June 12, 1978; Complaint Counsel's Answer To Motion Of Respondent, The Kroger Company, For Reconsideration Of The Order Of May 17, 1978, Or Alternatively, For An Interlocutory Appeal From That Order, dated June 15, 1978; Order Ruling On The Kroger Company s Motion For Reconsideration Of The Order Of May 17, 1978, Or, Alternatively, For An Interlocutory Appeal, dated June 19, 1978). 110. Kroger has represented in certain advertisements, directly or by implication, that the results of the Price Patrol Survey prove that shopping at Kroger, rather than at competitors' stores, will result in lower overall (34)expenditures. (Complaint Counsel's Motion For Summary Decision, dated February 24, 1978; Memorandum Of Respondent, The Kroger Company, In Opposition To Complaint Counsel' s Motion For Summary Decision, dated April 14, 1978; Reply To Memorandum Of Respondent, dated April 24, 1978; Reply Memcrandum Of Respondent, dated May 6, 1978; Order Ruling On Complaint Counsel's Motion For Summary Decision, dated May 17 1978 (pp. 4-6); Motion Of Respondent, The Kroger Company, For Reconsideration Of The Order Of May 17, 1978, Or, Alternatively, For An Interlocutory Appea, From That Order, dated June 12, 1978; Complaint Counsel's Answer To Motion Of Respondent, The Kroger Company, For Reconsideration Of The Order Of May 17, 1978, Or Alternatively, For An Interlocutory Appeal From That Order, dated June 15, 1978; Order Ruling On The Kroger Company s Motion For Reconsideration Of The Order Of May 17, 1978, Or, Alternatively, For An Interlocutory Appeal, dated June 19, 1978. 111. Kroger has represented in certain Price Patrol advertisements, directly or by implication, that shopping at Kroger, rather than at competitors' stores, will result in lower overall expenditures- (Complaint Counsel's Motion For Summary Decision, dated February 24, 1978; Memorandum Of Respondent, The Kroger Company, In 674 F.;DERAL TRADE COMMISSION DECISIONS Initial Dccision 98 F. Opposition To Complaint Counsel's Motion For Summary Decision dated April 14, 1978; Reply To Memorandum Of Respondent, dated April 24, 1978; Reply Memorandum Of Respondent, dated May 6 1978; Order Ruling On Complaint Counsel's Motion For Summary Decision, dated May 17, 1978 (pp. 4--); Motion Of Respondent, The Kroger Company, For Reconsideration Of The Order Of May 17 1978, Or, Alternatively, For An Interlocutory Appeal From That Order, dated June 12, 1978; Complaint Counsel's Answer To Motion Of Respondent, The Kroger Company, For Reconsideration Of The Order Of May 17 , 1978, Or, Alternatively, For An Interlocutory Appeal From That Order, dated June 15, 1978; Order Ruling On The Kroger Company s Motion For Reconsideration Of The Order Of May 17, 1978, Or, Alternatively, For An Interlocutory Appeal, dated June 19, 1978.

112. Kroger has represented in certain advertisements, directly or by implication, that the Price Patrol Survey is a methodologically sound survey. (Complaint Counsel's Motion For Summary Decision dated February 24, 1978; Memorandum Of Respondent. The Kroger Company, In Opposition To Complaint Counsel's Motion For Summary Decision, dated April 14, 1978; Reply To Memorandum Of Respondent, dated April 24, 1978; Reply Memorandum Of Respondent, dated May 6, 1978; Order (35JRuling On Complaint Counsel' Motion For Summary Decision, dated May 17, 1978 (pp. 4-6); Motion Of Respondent, The Kroger Company, For Reconsideration Of The Order Of May 17, 1978, Or, Alternatively, For An Interlocutory Appeal From That Order, dated June 12, 1978; Complaint Counsel's Answer To Motion Of Respondent, The Kroger Company, For Reconsideration Of The Order Of May 17, 1978, Or, Alternatively, For An Interlocutory Appeal From That Order, dated June 15, 1978; Order Ruling On The Kroger Company s Motion For Reconsideration Of The Order Of May 17, 1978, Or, Alternatively, For An Interlocutory Appeal, dated June 19, 1978.) 113. Kroger has represented by implication in certain Price Patrol advertisements that at the time Kroger represented that shopping at Kroger, rather than at competitors' stores, would result in lower overall expenditures, it possessed and relied upon a reasonable basis for that representation. 114. Kroger officials testified that the absence of meat and produce items was not disclosed in Price Patrol advertisements (Tr. 768-69 1661 167&-77 1817 2031 2101 2445; JX 2: Stec, p. 88; JX 1: Hoover, p. 55; JX 1: Pellin, p. 68). A review of the advertisements submitted by both complaint counsel and respondent confirms the Tile KROGFeR CO. 675 639 Initial Decision fact that the absence of meat and produce was not disclosed in the Price Patrol advertisements.

115. The use of the word "grocery" in Price Patrol advertisements does not constitute a disclosure of the exclusion of meat and produce, since "grocery" is commonly understood by consumers to mean items generally sold in a supermarket or grocery store including meat and produce (Tr. 1117- , also see "grocer" and grocery" in Webster s Third New International Dictionary of the English Language, Unabridged (1961)).

115A. Meat comprises 21 % of Kroger s sales, and produce 7% (F. 29). The sales of meat and produce are, therefore, significant (Tr. 2985). Both Kroger expert witnesses agree that the absence of meat and produce should have been disclosed in the Price Patrol advertisements (Tr. 2676, 270&-10, 2988-9).

116. Beginning in 1976, some Price Patrol newspaper advertisements included fine-print disclosures that fresh meat and produce items were not included in the Price Patrol (RX 595, 1128; Hoover Tr. 1676-77). The Price Patrol check lists which were posted in Kroger stores and viewed by (36)significant numbers of consumers did not list fresh meat and produce items. To the extent that some later Price Patrol advertisements disclosed the exclusion of meat and produce from the Price Patrol advertisements, they were made on the direction of Kroger s General Office because of a private suit in Chattannoga, Tennessee (Tr. 1677), and the present liigation involving the Federal Trade Commission (Tr. 1816-17; CXs 22, 23). After such disclosures were made in certain Price Patrol advertisements, there was no fall-off in Kroger s sales (Tr. 1870). Central Marketing Area 117. The following advertisements from the Central Marketing Area were received into evidence:

CXs 406-09, 411-414, 422-427, 435-39;

RXs 500(3-4), 505, 518 1111, 1112-1114.

118. None of the above advertisements for the Central Marketing Area contain a disclosure that meat and produce are excluded from the Price Patrol Survey.

119. Numerous advertisements imply that meat and produce are included in the Price Patrol Survey. For example, CXs 435 through 439 have as a banner headline: "Want to Save More Each Week on Your Food Budget?" Additionally, a picture of the Price Patrol women is captioned: "These Three Area Housewives Are Becoming , Initial Dccision 98 F. The Best Informed Experts On Food Prices In Fort Wayne. " (CXs 436-39). In CXs 422-427, the headlines generally state that the price Patrol Proves Kroger Is The Low Price Leader " and a consumer will "save more" this week at Kroger. Dallas Marketing Area 120. The following advertisements from the Dallas Marketing Area were received in evidence:

CXs 400-405;

RXs 595, 598-600, 609, 1123-1128. (37) 121. A review of the advertisements in evidence reveals that the exclusion of meat and produce was not disclosed in Price Patrol advertisements until September 27, 1976 (CXs 400-401 , 404-405; RXs 598-599, 609, 1123-1127). RX 595 Dallas Times-Herald dated September 27, 1976, and RX 1128 Dallas Morning News dated October 4, 1976, contain the following disclosure: Price Patrol checks only items which are totally or practically identical in each of the stores checked. Fresh meat, fruits and vegetables cannot be accurately compared due to variations in grades, sizes, trim, etc 122. Some advertisements dated after September 27, 1976 and October 4, 1976 do not disclose the absence of meat and produce from the Survey. In fact, CX 402 Dallas Morning News dated October 21 1976 (17 days after October 4 1976), implies that meat is included in the Price Patrol report. The banner headline includes a reference to the Price Patrol, and the statement: "Kroger Meat Prices are Down (CX 402). CX 403 Dallas Morning News dated November 4, 1976 also includes the banner headline referencing the Price Patrol in close proximity to the large typed statement Kroger Meat Prices are Down" (CX 403).

Delta Marketing Area 123. The following advertisements from the Delta Marketing Area were received into evidence:

CXs 428-434;

RXs 551-553, 554(1-2), 555-556, 560-563.

124. None of the above advertisements contain a disclosure that meat and produce are excluded from the Price Patrol Survey. . . , , , . . . THE KROGm, CO. 677 639 Initial Decision Southland Marketing Area 125. The following advertisements from the Southland Marketing Area were received into evidence:

CXs 415-421;

RXs 535 (536), 537 (538), 540-542, 546. (38) 126. None of the above advertisements contain a disclosure that meat and produce are excluded from the Price Patrol Survey. 127. RX 540 implies that meat and produce are included in the Price Patrol Survey. On RX 540-1 is the statement shopping Kroger will enable you to spend less for your food than at any other store." On RX 540-2 is the statement They check prices on over 150 items. Not the kind of items you buy once a year, but basics. Staples every housewife needs in her kitchen everyday. Staples includes meat and produce (Tr. 1818).

128. Certain advertisements containing. price claims made by other food retailers have been admitted into the record (RXs 801- 803, 813, 830, 837-838, 840, 842, 844; Tr. 3091-92). The record is silent as to whether the claims made in these exhibits were substantiated, and, if so, on what basis. Some Effects of the Price Patrol Program 129. The Price Patrol advertisements were run by Kroger in some areas for almost six years 1972 through 1978 (F. 1(5). Most other advertising programs at Kroger had a life of from three to four months (Tr. 2093-1). The primary reason for Kroger s discontinuance of the Price Patrol was the pendency of this proceeding (F. 107). 130. The Price Patrol program had a remarkably positive effect on Kroger s sales, profits, and market share. For example, in Atlanta, during the life of the Price Patrol program, sales inereased significantly, Kroger s market share doubled and net profit margin also inereased (Tr. 1870, 2159). Similarly, in the Dallas Marketing Area sales increased during the Price Patrol program (Tr. 770; CXs , 708-3). CX 15, a Kroger document entitled "The Price Patrol" (F. 27), states clearly on the cover The Price Patrol really works like nothing before or since. to increase sales, profits, and credibility . . . to reduce advertising and promotional costs." (CX 15-1; emphasis in original). The manual goes on to point out that "(tJhe use of this type program will produce long term gains in share of market and profitability" (CX 15-2). The manual also attaches a sales chart from the Atlanta area (CX 15-2, 15-19). The chart shows that Kroger sales l39)per month almost doubled from the twelfth period in 1972, before , . .

Initial Dccision 98 F. the Price Patrol started, to the twelfth period of 1974 (CX 15-19). Furthermore, an attachment to the written speech of Mr. Thomas given to the Vice Presidents of other Kroger Marketing Areas (Tr. 2093-S) shows that, by March 1974, only 16 months after the Price Patrol's introduction, Kroger sales and customer count increased substantially (CX 30-13). The Price Patrol manual also points out that, as the program matures and sales increase, the media cost, as a percentage of sales, will go down (CX 15-13). Mr. Stec confirmed that this phenomenon did indeed occur (Tr. 1836 in camera). The manual also states that sales in the Southland Marketing Area, other than Atlanta responded well and are following the same up swing as did Atlanta." (CX 15-13; see also CX 29). Mr. Thomas' discussion of the increase in net profits in Atlanta is supported by CX 15-14. The manual also supports Mr. Thomas' figures of increased market share: " it continues to grow and enables Kroger to increase its market share and keep it. . ." (CX 15-14, 15-15). 131. Certain Kroger officials, particularly in the Southland Marketing Area, testified regarding the effects of the Price Patrol on competitive activity (Tr. 1678, 1768-69, 2107-08). This testimony, however, does not square with the Kroger Price Patrol manual which states, under the heading "Competitive Activity None of the competition in the markets where the Price Patrol has been introduced have changed their merchandising style. If they were a high priced feature merchant they continued this type activity and may have heated up features at the Qut-set to try to off:'set the affects of the price patrol. If a competitor is a EDLP merchant they have made a few adjustments to the regular structure but have not met the roller program. (eX 15-15). 132. CX 16, entitled "Managing A Kroger Price Patrol Program is a document that was disseminated to the Kroger Marketing Areas as an information guide for the Price Patrol program. The document details how to put together, run, and manage a Price Patrol campaign (Tr. 760). The document discloses that the results of the Price Patrol program tend to build gradually, rather than immediately. And the effective life of the program is such that, instead of being "worn out" (40Jafter a few months or a year, it will continue to pile up credibility and results indefinitely (CX 16-3). In areas where the Price Patrol had been run, consumer research showed high levels of public awareness and credibility (CX 16-3). The research also indicated that the Price Patrol concept attracted younger families (CX 16-3).

133. Even though meat and produce were not included in most Price Patrol checks, sales and consumer research showed that a THE KROGER CO. 679 639 Initial Decision hallo (sic) effect" improved Kroger s meat and produce price image as well (eX 1&-3), and Kroger s experience indicated that crossshopping is reduced, as reflected in higher average transactions (CX 1&-3).

134. Mr. Charles Thomas, in his speech to other Kroger Marketing Area Vice Presidents, points out the credibility and believabilty of the Price Patrol Program and says that the "awareness levels for Price Patrol are extremely high." (CX 30-9). 135. Kroger s sales in Atlanta increased substantially during the Price Patrol, its market share doubled and net profis also increased (Tr. 2159). Based on an assessment ofthe costs of the Price Patrol in relation to the effects, Kroger was satisfied with the effectiveness of the program (Tr. 2166).

136. Campbell-Mithun, Inc. , an advertising agency, conducted consumer awareness research regarding the Price Patrol in the following cities on the following dates (CX 708): Initial Decision 98 F. Atlanta, Georgia August 1- , 1973 March 18-23, 1974 April 11- , 1976 August 2 , 1977 November 2-4, 1977 Memphis. Tennessee February 8-11 , 1976 Albany, Georgia First week, February 1975 Macon, Georgia February 2-4, 1974 First week, November 1974 Augusta, Georgia October 18-20, 1973 Nashville, Tennessee September 20- , 1974(41) Indianapolis, Indiana February 8-11 , 1976 Little Rock, Arkansas February 18-20, 1976 Carrollon, Georgia March 27- , 1976 Ada, Oklahoma August 18-22, 1977 Duncan, Oklahoma August 18- , 1977 Athens, Georgia October 11-17, 1975 Shreveport, Louisiana October 31 - November 5, Dallas, Texas October 30 - November 2 1975; August 2 , 1977 137. These reports show that, as of the date of the respective surveys for each city (CX 708):

(a) A majority of food shoppers were aware of the Price Patrol and identified it with Kroger (except for the Augusta, Georgia and Duncan, Oklahoma surveys).

(b) Awareness of the Price Patrol and identification of it with Kroger had increased in the following cities for the following periods: Dallas, Texas October 30, 1975 to August 7, 1977 Macon, Georgia February 2, 1974 to November 1974 In Atlanta the results were as follows:

()!.

THE KROGER CO. 681 639 Initial Decision Percentage of Food Shoppers A ware of Price Patrol and Identify it with Kroger Aug. Mar. Apr. Aug. Nov.

1973 1974 1976 1977 1977 All shoppers 55% 67% 78% 85% 83% (42) (c) Of those food shoppers both aware of the Price Patrol and identifying it with Kroger, between 8% and 30% stated that they check the Price Patrol list available in the store. Between 70% and 92% stated that they did not.

(d) Of those food shoppers both aware of the Price Patrol and identifying it with Kroger, the following percentage showed increased interest in shopping at Kroger as a result of the Price Patrol in the cities and on the dates listed: AT KROGER LSEWHERE SHOPPERS ATLANTA , GEORGIA August 1973 2U% 13'7" 18% March 1974 29% 10% 16% Apr;! 1976 29% 11% 14% August 1977 i'O RESULTS November 1977 NO Rl' SIJLTS MACON , GEORGIA February 1974 42'Y" NO RESULTS 25% November 1974 26'%, NO RESIJ.'/S 18'11, DALLAS, TEXAS Oct. NQV 1975 14% 15% 15% August 1977 30% (e) Kroger sales increased during the period of the Price Patrol in Dallas, Texas.

138. The above awareness surveys were each made while the Price Patrol was in operation in the respective cities. They do not project the effect of the Price Patrol to periods after the date of the report (CX 708).

IV. Kroger Did Not Have A Reasonable Basis For Certain Claims Made In The Price Patrol Advertisements Deficiencies of the Price Patrol Survey 139. Complaint counsel's expert witness, Dr. Kohout, and Mr. Oliver, respondent's expert witness on survey methodology, (43)both agree that there are recognized standards by which to assess the methodological soundness of a price survey (Tr. 1004- , 1064 2707). llllLt n - R? - .',4 Initial Dccision 98 YT. 'rhe judgment that a survey is sound is based upon the application of accepted sampling techniques, proper statistical analysis, and the reduction of bias and random error (TL 1004- , 1065, 2708). Acceptable survey procedures also dictate that the limitations of the design be clearly stated (Tr. 1064, 1066). A survey is methodologically unsound when opportunities for systematic error (bias) and random error are not eliminated to the extent practicable ('lr. 1005). Random error is due to chance, and can be reduced by increasing sample size (Tr. 1005). Systematic error, or bias, cannot be reduced by increasing sample size (Tr. 1005-06). Bias is a systematic tendency to deviate from the true population (Tr. 2708). The soundness of the Price Patrol Survey can be assessed by the degree to which bias was minimized (Tr. 2708). A survey must be evaluated in light of the claims made (Tr. 1005). As shown by the following Findings, the Price Patrol program was conducted in such a way that there were a number of systematic errors which significantly biased survey results. Dr. Kohout, complaint counsel's expert witness, testified that, when measured against acceptable survey procedures, the Price Patrol Survey was not methodologically sound (Tr. 1011). 1. Representativeness of Items Incladed in the Price Patrol Survey 140. Kroger s Price Patrol Survey excluded certain categories of items in Kroger stores, particularly meat, produce. house brands most private label products, most direct-store-delivered items, bakery, and delicatessen items (Tr. 1120 2708 2854-55; F. 93). 141. Dr. Kohout testified that the exclusion of certain categories from the Survey presented a problem of representativeness (Tr. 1012, 1028), and such a survey excluding categories would not support the claim that Kroger was lower on more food items (Tr. 1029). Dr. Kohout succinctly stated the proposition: "(IJf one is to generalize to a certain population, all segments of the population have to be represented in the sample." (Tr. 1013, 1119). Mr. John Kroening, Account Executive for Campbell-Mithun, recommended against generalizing to 500 items from the Price Patrol Survey based on 150 items (CX 600).

142. Dr. Kohout also testified that the exclusion of meat and produce from the Price Patrol Survey prevented the Survey from supporting the claim that Kroger offers prices (44)which would result in lower overall expenditures (Tr. 1030). Such a survey would not support a conclusion that the average consumer would save money by shopping at the store (Tr. 1030, 1054-55). THE KROGER CO. 683 639 Initial Decision 143. Mr. Oliver, respondent' s expert witness on survey methodology. conceded that, to get an assessment of price position across the whole range of food categories, one cannot exclude meat and produce unless Kroger s price relationship to its competitors on meat and produce was the same as its price relationship on groceries (Tr. 2708- 11). Dr. Benham, respondent's expert witness, also agreed that: " meat prices differ systematically from the prices of groceries, indeed the grocery Price Patrol items are not representative of the overall prices in the store." (Tr. 2987).

144. The price relationship between Kroger and its competitors on meat and produce did, in fact, differ systematically from Kroger price relationship with its competitors on groceries, according to Kroger s own internal price checks.

145. Of all the Burgoyne Price Checks conducted in cities during the course of Price Patrol advertising in the Delta, Dallas, Central and Southland KMAs, Kroger lost to (that is, Kroger had higher prices than) one or more competitors sixty (60) out of three hundred eight (308) in the Grocery category (19%); two hundred sixty-four (264) out of three hundred eight (308) in the Produce category (86%); and two hundred fifty-three (253) out of three hundred eight (3D8) in the Meat category (82%).

146. Of all the Burgoyne Price Checks conducted in Bloomington Indiana during the course of the Price Patrol advertising campaign Kroger Jost to one or more competitors three (.3) out of six (6) in the Grocery category, four (4) out of six (6) in the Produce category, and five (5) out of six (6) in the Meat category (CX 251). 147. Of all the Burgoyne Price Checks conducted in Ft. Wayne Indiana during the course of the Price Patrol advertising campaign Kroger lost to one or more competitors four (4) out of nine (9) in the Grocery category, seven (7) out of nine (9) in the Produce category, and eight (8) out of nine (9) in the Meat category (CX 252). 148. Of all the Burgoyne Price Checks conducted in Indianapolis Indiana during the course of the Price Patrol advertising campaign Kroger lost to one or more competitors five (5) out of twelve (12) in the Grocery category, six (6) (45)out of twelve (12) in the Produce category, and eleven (11) out of twelve (12) in the Meat category (CX 253).

149. Of all the Burgoyne Price Checks conducted in Kokomo Indiana during the course of the Price Patrol advertising campaign Kroger Jost to one or more competitors three (3) out of six (6) in the Grocery category, six (6) out of six (6) in the Produce category, and five (5) out of six (6) in the Meat category (CX 254). 150. Of all the Burgoyne Price Checks conducted in Dallas, Texas 684 EDERAL TRADE COMMISSION DECISIONS Initial Decision 98 F. during the course of the Price Patrol advertising campaign, Kroger lost to one or more competitors six (6) out of twenty-one (21) in the Grocery category, twenty-one (21) out of twenty-one (21) in the Produce category, and fourteen (14) out of twenty-one (21) in the Meat category (CX 256).

151. Of all the Burgoyne Price Checks conducted in Sherman/Denison, Texas during the course of the Price Patrol advertis iog campaign, Kroger lost to one or more competitors one (1) out of eleven (11) in the Grocery category, ten (10) out of eleven (11) in the Produce category, and four (4) out of eleven (11) in the Meat category (CX 257).

152. Of all the Burgoyne Price Checks conducted in Ft. Worth Texas during the course of the Price Patrol advertising campaign Kroger lost to one or more competitors none of the eleven (11) in the Grocery category, eleven (11) out of eleven (11) in the Produce category, and seven (7) out of eleven (11) in the Meat category (CX 258).

153. Of all the Burgoyne Price Checks conducted in Shreveport Louisiana during the course of the Price Patrol advertising campaign, Kroger lost to one or more competitors one (1) out of twelve (12) in the Grocery category, ten (10) out of twelve (12) in the Produce category, and twelve (12) out of twelve (12) in the Meat category (CX 259).

154. Of all the Burgoyne Price Checks conducted in Ft. Smith Arkansas during the course of the Price Patrol advertising campaign, Kroger lost to one or more competitors one (1) out of nine (9) in the Grocery category, nine (9) out of nine (9) in the Produce category, and five (5) out of nine (9) in the Meat category (CX 260). 15,5. Of all the Burgoyne Price Checks conducted in Jackson Mississippi during the course of the Price Patrol advertising campaign, Kroger lost to one or more competitors (46Jnone of the six (6) in the Grocery category, six (6) out of six (6) in the Produce category, and four (4) out of six (6) in the Meat category (CX 261). 156. Of all the Burgoyne Price Checks conducted in Little Rock Arkansas during the course of the Price Patrol advertising campaign, Kroger lost to one or more competitors ten (10) out of thirteen (13) in the Grocery category, twelve (12) out of thirteen (13) in the Produce category, and ten (10) out of thirteen (13) in the Meat category (CX 262).

157. Of all the Burgoyne Price Checks conducted in Memphis Tennessee during the course of the Price Patrol advertising campaign, Kroger lost to one or more competitors none of the nine (9) in THE Kt(Uli tl vV.

639 Initial Decision the Grocery category, eight (8) out of nine (9) in the Produce category, and nine (9) out of nine (9) in the Meat category (CX 263). 158. Of all the Burgoyne Price Checks conducted in Pine Bluff Arkansas during the course of the Price Patrol advertising campaign, Kroger lost to one or more competitors none of the nine (9) in the Grocery category, eight (8) out of nine (9) in the Produce category, and seven (7) out of nine (9) in the Meat category (CX 264). 159. Of all the Burgoyne Price Checks conducted in Atlanta, Georgia during the course of the Price Patrol advertising earn paign, Kroger lost to one or more competitors eleven (11) out of twenty-five (25) in the Grocery category, seventeen (17) out of twenty-five (25) in the Produce category, and twenty (20) out of twenty-five (25) in the Meat category (CX 274).

160. Of all the Burgoyne Price Checks conducted in Augusta, Georgia during the course of the Price Patrol advertising campaign Kroger lost to one or more competitors three (3) out of sixteen (16) in the Grocery category, fifteen (15) out of sixteen (16) in the Produce category, and thirteen (13) of fifteen (15) in the Meat category (CX 275).

161. Of all the Burgoyne Price Checks conducted in Bowling Green (Southland KMA) during the course of the Price Patrol advertising campaign, Kroger lost to one or more competitors none of the thirteen (13) in the Grocery category, thirteen (13) out of thirteen (13) in the Produce category, and thirteen (13) out of thirteen (13) in the Meat category (CX 276). 162. Of all the Burgoyne Price Checks conducted in Chattanooga Tennessee during the course of the Price (47JPatrol advertising campaign, Kroger lost to one or more competitors three (3) out of twenty-one (21) in the Grocery category, nineteen (19) out of twentyone (21) in the Produce category, and nineteen (19) out of twenty-one (21) in the Meat category (CX 277).

163. Of all the Burgoyne Price Checks conducted in Florence/Sheffield, Louisiana during the course of the Price Patrol advertising campaign, Kroger lost to one or more competitors two (2) out of twelve (12) in the Grocery category, eight (8) out of twelve (12) in the Produce category, and eight (8) out twelve (12) in the Meat category (CX 278).

164. Of all the Burgoyne Price Checks conducted in Huntsvile Alabama during the course of the Price Patrol advertising campaign Kroger lost to one or more competitors none of the thirteen (13) in the Grocery category, seven (7) out of thirteen (13) in the Produce category, and nine (9) out of thirteen (13) in the Meat category (CX 279).

Initial Decision 98 F. 165. Of all the Burgoyne Price Checks conducted in Knoxville, Tennessee during the course of the Price Patrol advertising campaign, Kroger lost to one or more competitors two (2) out of thirtyseven (37) in the Grocery category, thirty-five (35) out of thirty-seven (37) in the Produce category, and thirty-six (36) out of thirty-seven (37) in the Meat category (CX 280).

166. Of all the Burgoyne Price Checks conducted in Macon, Georgia during the course of the Price Patrol advertising campaign Kroger lost to one or more competitors none of the fifteen (15) in the Grocery category, eleven (11) out of fifteen (15) in the Produce category, and fourteen (14) out of fifteen (15) in the Meat category (eX 281).

167. Of all the Burgoyne Price Checks conducted in Nashvile, Tennessee during the course of the Price Patrol advertising campaign, Kroger lost to one or more competitors six (6) out of twentytwo (22) in the Grocery category, twenty-one (21) out of twenty-two (22) in the Produce category, and twenty-one (21) out of twenty-two (22) in the Meat category (CX 282).

168. Mr. Oliver sponsored respondent's exhibits RXs 920-924 (Tr. 2609- , 2634). These exhibits purport to show that "Price Patrol items were fairly well distributed throughout the grocery department of Kroger." (48)(Tr. 2634). These exhibits, of course, exclude meat, produce, and delicatessen items (Tr. 2745-6). These exhibits do not show the representativeness of the Price Patrol item as a percentage of sales throughout the grocery department (Tr. 2749). Furthermore, Mr. Oliver combined Price Patrol lists for his analysis (Tr. 2749) and RXs 920-924 do not show how representative the Price Patrol list was for the one week it was checked (Tr. 2749). 169. The selection of a survey sample by informed judgment and expertise is a short cut generally recognized in the field of survey design (Oliver, Tr. 2660-66; Benham, Tr. 2827, 2832, 2841--4, 2847- , 2860-4; Kohout, Tr. 1154-58). Judgment samples are samples drawn on the basis of informed judgment and expertise rather than random sample methodology (Oliver, Tr. 2660-61; Kohout, Tr. 1301 1314). As such, the findings derived from data obtained through the use of judgment sample is not projectible to any general population. However, the results of judgment sample surveys may provide useful insights into true facts or useful broad guides to decision making (Tr. 2661-64 2827 2838 2841--4 2847- 2660- 3083). 170. The results of judgment samples are commonly used by business, labor, government agencies, and researchers as convenient aids to decision making (Oliver, Tr. 2666; Benham, Tr. 282&-27, 2832 2841--4 2847- 2860-64).

THE KROGER CO. 687 6:19 Initial Decision 171. Judgment sampling methodology seeks to bring the accumulated knowledge and insight of experts to bear in constructing a survey sample and in evaluating the resulting survey data (Oliver Tr. 2668, 2695- , 2715-16).

172. Another aspect of judgment sampling is that, depending upon the purpose in view, it may resolve problems which prove difficult when random sampling methodology is employed (Oliver Tr. 2647- , 2782-83; Benham, Tr. 2889, 2893-94). 173. Another aspect of judgment sampling is that the costs are substantially less than the costs of random sample methodology (Oliver, Tr. 2641-42; Benham, Tr. 2889, 2893-94). 174. Experienced survey researchers often must give careful consideration to tradeoffs among practicality, feasibility, cost, and the expected quality and importance of the information produced by alternative survey methodologies. having in mind the use to which the survey is to be put (Benham, Tr. 2885-95). (49) 175. It is the opinion of respondent's experts that judgment sample methodology is useful for conducting retail food price surveys (Oliver, Tr. 2647-49, 2715; Benham, Tr. 2863-64) and that the accumulated knowledge and insight of persons experienced in food retailing can be brought to bear on constructing a judgment sample and evaluating its results (Oliver, Tr. 2668, 2670, 2715-16). 176. According to Mr. Oliver, the use of judgment samples for conducting retail food price surveys avoids the difficult methodological problems caused by quality differences among items between stores and nonavailability of comparable items between stores and is reasonable under the circumstances (Oliver, Tr. 2643, 2645, 2647- 2782-83).

177. The Price Patrol Survey is a judgment sample survey and as such, could have provided Kroger a convenient and inexpensive shorthand device for obtaining further clue to its true competitive position, when evaluated in conjunction with Kroger s other price checks, including Full Book Checks and Burgoyne Checks. 2. Failure to State Limitations of the Price Patrol Survey Design 178. All three experts testifying on survey methodology were unanimous that the systematic exclusion of fresh meat and produce from the Price Patrol Survey should have been stated as a limitation in the survey design in all Price Patrol advertisements (Tr. 1066 2709 2988-89).

Initial Decision 98 FTC. Systematic Inclusion of'Reduced-Price Items in the Price Patrol Survey 179. The Price Patrol Survey had a far greater representation of roller items in the store. Additionally, the merchandiser selecting the Price Patrol items was aware of which items were rollers. The criteria were the same for selecting items as rollers and for selecting items to be included on the Price Patrol Survey ('fr. 1675-76; F. 34 , supra). One grocery merchandiser regularly placed 20 to 40 items on the Price Patrol list precisely because they were rollers (F. 95 supra).

180. Dr. Kohout testified that a systematic inclusion of roller items on the Price Patrol lists presents a bias and reasonable steps should have been taken to eliminate (50)that bias (Tr. 1033). Dr. Kohout explained that there can be bias in the procedures or in the results of a survey. Bias in procedures may not always show up in the results ('fr. 1143). He further testified that knowledge of roller items and inclusion of rollers on the Price Patrol is a source of bias that should have been eliminated, even if the rollers were not overrepresented (Tr. 1035, 1039). Dr. Kohout suggested that, if the selection of Price Patrol items were done randomly or by an outside source, this bias might be eliminated (Tr. 1037). "But someone who has a personal interest in the outcome, a personal interest in supporting the hypothesis, putting them on, it' s just inherently wrong. " (Tr. 1037).

181. Dr. Kohout further testified that the over-representation of rollers on the Price Patrol inflates the result of the estimate of the number of items that Kroger is lower on (Tr. 1034), and the Price Patrol would, therefore, not be representative of the total population (Tr. 1035). Mr. Oliver, Kroger s own witness, agreed that one expects rollers to be over-represented on the Price Patrol, and that such over-representation biases the results in favor of Kroger (Tr. 2735). The Price Patrol Survey was biased with respect to roller items because of the manner in which those items were placed on the Price Patrol Survey lists, and also because those items were over-represented in relation to their actual proportion in the population (Tr. Jj:30- , 1135, 1137 39).

(Joe of High- Volume Items 182. Kroger selects high-volume, fast-moving items for the Price Patrol. Dr. Kohout testified that this procedure is unsound (Tr. 1011 , 1039) because the fast-moving items comprise a minority of the Grocery category in terms of sales (Tr. 1039-43; ex 803). RX 979 lrID f'nVUI".dl. \.A.1. uoo 639 Initial Decision supports Dr. Kohout's proposition that Price Patrol items represent a minority of grocery items (18.9%), and only 11.2% of total store sales (RX 979).

Grocery Merchandisers who Selected Price Patrol Items Also Priced Items 183. Grocery merchandisers who selected items for the Price Patrol lists (F. 64; Tr. 2713) also had the responsibility for pricing those same items (F. 33; Tr. 2713). (51) 183A. The instruction manual for conducting the Price Patrol in the Delta Marketing Area states, under a general heading, t'Mechanics of Handling Price Patrol"

Price Patrol Checks.

3. Grocery Merchandising wil be responsible for being favorably competitive in all price checks. Individual store managers will be responsible for correct pricing at store leuel according to the latest information from Grocery Merchandi.',ing. (CX 75- 13. Emphasis in orig-inal), 184. Dr. Kohout testified that combining the dual functions of pricing and selecting items to be included in the survey increase the potential for biased item selection (Tr. 1043-44), and the procedure is unsound, based on accepted principles or survey methodology (Tr. 1011-12). Dr. Kohout would not change his opinion, even if the grocery merchandiser did not know which actual Price Patrol list was to be used in any given week (Tr. 1046). Both of respondent' expert witnesses Mr. Oliver and Dr. Benham agreed that bias was introduced by permitting the person to price items and select items for the Price Patrol Survey, and would have recommended separating these functions (Tr. 2713, 2989). Dr. Benham noted that it is important to separate these functions since bias may be introduced consciously or unconsciously (Tr. 2989-90). The Price Patrol Survey Sample Was Not A Representative Sample 185. The grocery merchandiser did not select the items for incl usion in the Price Patrol on a random basis (F. 64- supra; Tr. 2712).

186. The standard procedure for drawing a sample is to draw as many as possible in a neutral fashion, or to do so randomly so that there would be some assurance that the selection is done blindly (Tr. 1043). If one has a stake in the survey outcome, the usual procedure is either random sampling, or somehow to make that selection blind Initial Decision 98 F. neutral (Tr. 1043--4). If the sample is not selected randomly, it must be chosen by an independent third party who has no stake in the outcome of the survey ('Ir. 1066 2715-16). (52) 187. Random sampling makes it possible to make judgments about the efficiency or precision of the sample. Dr. Kohout testified that the absence of a random selection meant that each item in the store did not have a proportionate chance of being selected on the Price Patrol (Tr. 1048), and that, therefore, generalization based on probability theory cannot be made from such a survey (Tr. 1047--8). Presentation Format Used in the Price Patrol Advertisements 188. The Price Patrol Surveys were recapped and advertised in the form of a box score listing the number of items Kroger was higher on, lower than, or the same on price as its listed competitors. This box score did not list the prices of the items, nor did the chart summarize the total dollar costs of the items included. 189. Dr. Kohout testified that presenting the Survey results viewed in terms of the number of lower, number of higher, and number of same would not support any conclusion about whether or not consumers would save money at that store (Tr. 1051-52). Dr. Kohout points out that it is possible to have higher prices overall and still come out with more lower-priced items on the Survey (Tr. 1052). 190. Mr. Dean Hicks, who worked in the grocery business for 27 years, stated in discussing Full Book Recaps (CX 220) that, based on the higher-than, lower-than, or same as format, even he could not tell which store is the cheapest place to shop (Tr. 2435). Mr. Oliver respondent's expert witness, stated that consumers would be primarily interested in the amount of money which would be saved, and would find price information more useful than the number of items priced lower at a given store (Tr. 2766, 2770). Respondent prepared an exhibit costing out the items on the Price Patrol Survey for the week ending January 14, 1978 in Dallas (RX 965). Respondent does not contend that this exhibit is representative of other checks on other dates in other cities (Tr. 2765). Mr. Oliver admitted that the box score item results overstated the dollar differences between Kroger and its competitors (Tr. 2766). For example, the item count in RX 965 showed Kroger winning by a 4 to 1 ratio, while the dollar difference was only 1.01 to 1 (Tr. 2769-70). (53) Weighting 191. The Price Patrol Survey was not weighted (Tr. 2730-31). 192. Weighting" is a process of statistically adjusting the THE KROGER CO. 691 639 Initial DecisiOh sample to make it more representative (Tr. 1053). It is possible to design a self-weighted sample by assigning differential probabilities to each item in the sample population. Dr. Kohout testified that weighting involves statistically adjusting the sample to make it more representative (Tr. 1053). This weighting is necessary to make the Price Patrol Survey methodologically sound (Tr. 1012, 1052-53). 193. Factors necessary for weighting a survey which is designed to answer the question of which competitor offered prices resulting in lower overall expenditures, are the price and the average consumer expenditures by category of item (Tr. 1053, 2732- , 2990). Basically, Dr. Kohout would recommend weighting the items to take into account the frequency of purchase, and possibly the proportion spent in various categories (Tr. 1053-54). Mr. Oliver and Dr. Benham concurred with the proposition that weighting would have improved the reliability of the Price Patrol Survey (Tr. 2731 2992). 9. The Price Patrol Survey Did Not Support The Generalizations in Price Patrol Advertisements 194. Conclusions of the Price Patrol Survey could not be generalized beyond the 150 items actually checked in the Survey err. 1146). Furthermore, any claim limited to the 150 items surveyed must disclose that these items constitute a hand-picked sample (Tr. 1148- 49).

195. Dr. Kohout testified further that the various deficiencies in the Price Patrol Survey lead to the concl us ion that the Price Patrol Survey was not a methodologically sound survey (Tr. 1011- , 1054 1119).

196. Even if the only deficiency of the Price Patrol were the exclusion of meat and produce, the survey could not answer the question of whether Kroger offered prices which resulted in lower overall expenditures (Tr. 1084-85), nor could it answer the question of whether most items in Kroger stores are lower than items in competitors' stores (54)(Tr. 1085-86). "If we exclude the meat and produce categories, assuming that all other sources of bias could be controlled, we re stil generalizing beyond what we could support" (Tr. 1085).

197. Even if the only deficiency of the Price Patrol Survey were the fact that the criteria for the selection of Price Patrol items was the same as the criteria for selection of rollers, the Price Patrol Survey could not answer the question of which store offered prices which would result in lower overall expenditures (Tr. 1087-88). 198. Dr. Kohout also stated in summary that the Price Patrol , pp.

Initial Decision 98 F. Survey could not support a conclusion showing which competitor had prices which resulted in lower overall expenditures (Tr. 1054-55), nor could the Price Patrol support a conclusion that most items in Kroger stores are priced lower than in competitors' stores (Tr. 1057- 58).

10. Indiana Public Interest Group (INPIRG) Study 199. The Indiana Public Interest Group (INPIRG) began a study of the Price Patrol program in Indianapolis during the fall of 1975 in order to determine the reliability of the Price Patrol program (Tr. 455). INPIRG focused on price increases and decreases before an item appeared on the Survey (Tr. 456-57). In analyzing the raw data gathered by INPIRG, Professor Kleyle, an expert in statistical techniques, performed a Chi-square analysis of the data (Tr. 588-89). The value of the Chi-square indicates whether there is an association between the observed and the expected values. In the case of the Price Patrol, there were fewer Price Patrol items showing price increases than one would expect (Tr. 604). The number of items on the Price Patrol showing price decreases was larger than expected. And the high Chi-square results indicate that the likelihood of these changes being due to chance alone is extremely small (Tr. 593-96). Dr. Kleyle concluded that, because of the high Chi-square values, it is extremely unlikely that these variables of classification, that is whether an item was on a current Price Patrol list and whether the price of the item increased or decreased from the previous week (Tr. 594), are not associated. (55) Kroger Did Not Rely on its Other Price Checks As Substantiatiun for Claims Made in Price Patrol Advertisements 1. Full Book Checks 200. Kroger Marketing Area price-checking personnel conduct a price survey of approximately 2 000 to 4 000 items in a Kroger store and the stores of certain competitors (Tr. 2345--6; JX 2: Hoover 151 , 153; JX 1: Crague, p. 46; JX 1: Hoover, pp. 28, 30). This check is commonly referred to as a Full Book Check (Tr. 775, 797, 1564, 2002; JX 1: Pellin, pp. 73, 75-76; JX 1: Crague, p. 44). The Full Book Check results were summarized under the following headings: Grocery, Dairy, Frozen Food, DSD (direct store delivered items), H&BA (health and beauty aids), Total Grocery, Top 200 (or Top 500), and occasionally Produce, Meat (CXs 201- , 220-26; see also, RXs 947- 964).

g., THE KROGER CO. 693 639 Initial Decision 201. The Full Book Check is generally conducted at least once a quarter (Tr. 2112, 2342), and, in some areas, every four to five weeks (Tr. 1570, 1653; JX 2: Hoover, p. 151). The results of the Full Book Check are summarized in the format "higher than lower than " or the same as (e. CX 201; Tr. 1929-30), which is the same format used in the Price Patrol summaries (e. CX 125; JX 1: Pelln, p. 76). The Full Book Check (CXs 24- , 25-3) must be submitted once a quarter to Kroger s General Offce (Tr. 1570, 1613, 2100, 2342, 2348). CXs 209 and 210 are typical examples of Full Book Recaps submitted to Kroger s General Office (Tr. 2344).

202. The Full Book survey checked major Kroger competitors (JX 2: Hoover, pp. 151-52; Tr. 1565, 2345), and was used to determine Kroger s competitive price position in the market (Tr. 1564, 1570; JX 1: Pellin, p. 75; Tr. 2132, 2137, 2407).

203. The Full Book Price Checks sometimes include meat (Tr. 1927, 2346; JX 1: Crague, p. 50), produce (Tr. 2003, 2346; JX 1: Crague, p. 50), and private label items (Tr. 2347). Mr. Vaughn testified that approximately 30 to 50 fresh meat items are checked on the Full Book (Tr. 1928). He said that they attempt to include the basic meat cuts from each major meat commodity, including such items as chuck steak, ground beef, rib eye steak, quarter pork loins and center chops (Tr. 1928). The Kroger meat merchandiser or his merchandising representatives does the actual checking (56J(Tr. 1928). Dr. Bere testified that between 50 and 80 produce items are checked on the Full Book by either the Kroger produce merchandiser or his merchandising representative (Tr. 2003-04, 2032). 204. During the course of the Price Patrol advertising campaign Full Book Checks were conducted in certain cities in the Central Dallas, Delta, and Southland Marketing Areas (CX 200; RX 978). 205. Full Book results are not regularly compared with Price Patrol results (JX 1: Hoover, pp. 30, 46; JX 2: Hoover, p. 184). 206. Key Kroger officials testified that the Full Book Checks were not relied upon as substantiation for any Kroger advertising (Tr. 783-84, 1656, 1671, 1823-26; JX 1: Hoover, pp. 30, 39, 46; JX 2: Hoover, p. 184). Mr. Helmsing testified that Full Book Checks have nothing to do with Price Patrol advertisements ('lr. 783-84). Mr. Hoover stated:

Q. My understanding is that the Full Book was not used in any way in the advertising. Is that correct? A. That's correct.

Q. My understanding is also that you did not rely on the Ful) Book for your advertising. Isn t that correct? A. That is correct. (Tr. 1656) , pp.

Initial Decision 98 F. Mr. Stec was asked the following question: Were the advertising claims based in any way upon the full-book checks? . . (1'r. 1823) He answered:

, they were not based on the full-book checks. (Tr. 1823) 207. The Full Book Checks could not, in any event, have been used to substantiate the weekly Price Patrol claims (Tr. 1653; F. 208- 214, infra).

208. The Price Patrol checks were conducted weekly (CX 5-3), while the Full Book Checks were conducted only (57Jonce every four to five weeks (Tr. 1653, 2166-7, 2169). Mr. Vaughn and Mr. Hoover testified that meat and grocery prices change rapidly (Tr. 1659 1947), and Mr. Thom2s testified that the only weekly price check that was conducted by Kroger on a regular basis was the Price Patrol Survey (Tr. 2169). Dr. Kohout correctly noted that Full Book Checks conducted every three or four weeks could not support a weekly price claim (Tr. 1059-60). Kroger s own documents recognize that Price Patrol advertising claims could only be supported by weekly price checks (CXs 16-5, 73-5).

209. The Full Book Checks could not have been relied upon as substantiation for certain Price Patrol advertisements because, in certain cities where the Price Patrol program was conducted, no Full Book Checks were made (Tr. 2429). For example, RX 978 shows that ull Book Checks were not made in cities such as Kokomo, Pine Bluff, Florence/Sheffield, Bowling Green, and Duncan (RX 978 2). These same cities ran the Price Patrol program (RX 900). CX 814 shows that only the Price Patrol, and no other check, was conducted in 43 cities in the Delta Marketing Area (Tr. 1332-33). Mr. Oliver, Kroger s expert witness, confirmed that Full Book Checks were not conducted in 20% of the areas where the Price Patrol was run (Tr. 2786).

210. The Full Book Checks could not have been relied upon as substantiation for certain Price Patrol advertisements, because, in certain cities where both checks were conducted, the competitors checked differed. Mr. Pellin stated that the same competitors are not checked on both checks (JX 1: Pellin, pp. 75-76). Mr. Helmsing pointed out that eight competitors were checked on the Price Patrol in Dallas (CX 401), while only five competitors were checked on the corresponding Full Book Check (CX 204). From at least November , 1975 to August 23, 1976, Kroger checked Treasure Island on its Full Book Checks in Atlanta, Georgia (CX 220-12 to 220-15; Tr. g., THE KROGER CO. 695 639 Initial Decision 2430-32). For the corresponding period of Price Patrol Surveys Treasure Island was not checked on any Price Patrol (Tr. 2433). Treasure Island beat Kroger on three of the four Full Book Checks (CX 220-12, 220- , 220-15).

211. In Atlanta, Georgia, the Price Patrol Survey for the week ending December 20, 1975, indicated that eight competitors were checked by the Price Patrol (CX 415). The closest Full Book Check however, from November 24, 1975, only checked three competitors (CX 220-12). The Price Patrol advertisement for December 20, 1975 contained in part, (58)the following message: " 162 WEEKS OF CHECKING. . . SHOWS KROGER IS THE LOW PRICE LEADER" (CX 415), when Kroger s Full Book Check (CX 220-12) shows that Treasure Island beat Kroger (had more lower prices (Tr. 2141-44 2431-32)) in the Grocery and Total Grocery categories. 212. In Indianapolis, Indiana on January 28, 1976, the Price Patrol advertisement indicated that a total of seven competitors were checked by the Price Patrol (CX 422), while the closest Full Book Check for that period, January 5, 1976, showed that only five competitors were checked (CX 201-3). The Price Patrol advertisement (CX 423) contained, in part, the following message: " Straight Weeks The Low Price Leader Proven Week After Week by The Price Patrol " when the closest Full Book Check, dated within four weeks of the Price Patrol Survey, showed that Marsh beat Kroger in both the Grocery and Total Grocery categories (CX 201-3). 213. The Full Book Checks could not have been relied upon as substantiation for certain Price Patrol advertisements because certain Full Book results were inconsistent with the Price Patrol claims. Mr. Thomas testified that, from December 1972 until early 1978, each weekly Price Patrol Survey result in Atlanta showed Kroger winning the Price Patrol (Tr. 2140). Mr. Thomas also testified that Kroger lost Full Book Checks in Atlanta during this same period of time (Tr. 2141-42). RX 947 shows that fourteen (14) Full Book Checks were made in Atlanta from 1973 through 1976 (RX 947- 947-2). Of these fourteen (14) Full Book Checks, Kroger lost to one or more competitors eleven (11) times (RX 947- , 947-2). Even assuming that Kroger exhibits RXs 947-964 (Summary of Full Book Survey Results) generally agreed with Price Patrol results, these exhibits could not be used to support the representations made in Kroger s Price Patrol advertisements, since the exhibits show that frequently meat and produce were not included on the Full Book Checks (RXs 947-964. See RXs 947-4, 947-5 and 948-4, 948-5). CX 813 shows numerous instances where the Full Book results were inconsistent with Price Patrol claims (Tr. 1342-55). g., Initial Decision 98 FT. 214. Dr. Kohout was of the opinion that Kroger Full Book Checks would not support a conclusion as to which competitor had prices which would result in lower overall expenditures (Tr. 1057-58). His conclusion was based on the fact that the format-higher than, lower than, same as (e. CX 200-was not in the form of prices (Tr. (59) 1057-59), i. not in dollars and cents (Tr. 1059). Dr. Kohout further pointed out that Full Book Checks, conducted every three or four weeks, could not support a weekly claim (Tr. 1059-60). Dr. Kohout also testified that a Full Book Price Check would not support a claim of Kroger having more lower prices than its competitors (Tr. 1060), unless the competitors checked on the Full Book Check were chosen randomly or represented a full census of the competitor population (Tr. 1060-1).

Burgoyne Checks 215. Burgoyne, Inc. is a marketing research firm headquartered in Cincinnati, Ohio, which was employed by Kroger to conduct price checks in certain cities (JX 2: Ticeson, p. 61; Tr. 2321). Kroger wanted an independent agency to make the checks (Tr. 2320). The Burgoyne Price Checks were started in January of 1973 (Tr. 2322- 23). Presently, the checks are conducted in over one hundred cities (Tr. 2323). The Burgoyne Check is frequently referred to as a market basket- or a shopping basket-type of check (Tr. 2320, 2323). The individual checks are generally conducted once a quarter or four times per year (Tr. 2H2, 2323), and are scheduled by the Kroger Central Office (Tr. 1571- , 2323). The individual KMAs did not know in advance when the Burgoyne Check was to he conducted in its area (Tr. 2323). The Kroger Central Office also selects which stores will be surveyed; a Kroger store and certain competitors stores (Tr. 2323-24). Kroger picked the items to be surveyed (Tr. 2326; JX 1: Crague, p. 51). Competitors checked include the price leaders as identified by the various KMAs (Tr. 2324). 216. The Burgoyne Check consists essentially of three parts: grocery, produce, and meat (Tr. 1958, 2002, 2041, 2325). There is a further breakdown for health and beauty aids within the grocery check (Tr. 2325), and the items tend to be national brands and certain commodity-type items (Tr. 2325). The category titled "Total" on the Burgoyne recap is computed by summing the retail prices from the other categories (Tr. 2329). Grocery items were selected to be representative of the product mix on the grocery commodity analysis (Tr. 2326). About thirty items are included in the produce section of the check, and these items are selected by the produce Tile KROGEH CO. 697 639 Initial Decision merchandising department at the General Office (Tr. 2330). These items are also selected to be representative of the sales for each of the major categories within the produce department (Tr. 2330). (60) Certain of the produce items are weighted to make that item representative of the category sales within produce (Tr. 2330). Approximately forty items are included on the meat portion of the check (Tr. 2332). These items are selected by Kroger s Central Office Meat Merchandising Department, and the items are weighted to have a proportionate relationship to the Kroger meat product mix (Tr. 2332-33). Kroger s ability to properly weight the Burgoyne Check is shown by RX 967, wherein respondent compared the actual weighting of the Burgoyne Check to the weights used for the Consumer Price Index, December 1977 (RX 967, p. 2). The results are substantially similar (RX 967).

217. The Burgoyne price checkers are either Burgoyne employees or agents of Burgoyne (Tr. 2325).

218. After the checks are made and returned to Kroger, the prices of grocery items are totaled (Tr. 2329-30), the prices of produce items are totaled, and then factored to represent 7%-8% of total store sales (Tr. 2332), and the prices of meat items are totaled and then factored to represent approximately 20% of the total store sales err. 2134, 2334). The totals are then transferred to a recap sheet (Tr. 2334). HX 975 is a typical Burgoyne Check, and page 1 of that exhihit is the recap sheet (Tr. 2321-22). 219. Mr. Charles Thomas explained how he analyzed the results of a typical Burgoyne recap sheet (CX 274-8; 'lr. 2133). For example Mr. Thomas looked at CX 274- , and from the face of that recapped check, he concluded that Kroger lost in the Grocery category to Treasure Island (Tr. 2135). A loss on a Burgoyne Check occurs when Kroger shows a larger total dollar amount on the face of a recapped Burgoyne Check (Tr. 2135).

220. The Burgoyne Checks are made to verify Kroger s competitive posture in each of the areas where the checks are conducted (JX 2: Ticeson, p. 62; JX 2: Ticeson, p. 74; JX 2: Stec, p. 97; Tr. 1960 2132). The results of the Burgoyne Check, however, are not disclosed to consumers (Tr. 2170). Burgoyne Checks are not regularly seen by the advertising managers in the various KMAs (JX 1: Stec, p. 92; JX 1: McDaniels, p. 24).

221. Burgoyne Checks are not used for advertising (JX 1: Hoover p. 63; JX 1: Pellin, p. 82; JX 2: Ticeson, p. 68; Tr. 1614, 182:J). The Burgoyne Check results am not compared with the Price Patrol results (JX 2: 'ficeson, pp. 68-69; Tr. 1612, 1672; see, CX 61). (61) 222. Key Kroger officials testified that the Burgoyne Checks 44 'J n - R? - J, 1), .

, p.

Initial Decision 98 F. were not relied upon as substantiation for any of Kroger s advertising claims (JX 2: Ticeson, pp. 68-69; JX 1: Stec, pp. 91-92; JX 1: Hoover, p. 63; JX 2: Stec, p. 107).

223. Mr. Helmsing testified that only the Price Patrol Surveys were relied upon as substantiation for Price Patrol advertisements in the Central and Dallas Marketing Areas (Tr. 783). Mr. Hoover and Mr. Stec confirmed that the Burgoyne Checks were not relied upon as substantiation for Price Patrol advertising in the Southland Marketing Area (Tr. 1672- , 1823; JX 1: Hoover, p. 39; JX 1: Stec 92; JX 2: Stec, p. 107). Mr. David Ticeson/Tyson, who was the Kroger official reponsible for the rescheduling and the coordination and tabulation of the Burgoyne Checks (JX 2: Ticeson, p. 60), states: Q. Okay. Those checks have nothing to do with the Kroger advertising program do they? A. No, sir, they do not. (JX 2: Tieeson, p. 68). Q. Are they used to verify whether the Price Patrol reports are getting accurate information on their surveys? A. No, sir, they are not. (JX 2: Ticeson, p. 68). Mr. Dean Hicks, Kroger s Vice President of Marketing Administration (Tr. 2306) with primary responsibility for the Burgoyne Checks at Kroger (Tr. 2426), confirmed that the Burgoyne Check was not actually used by Kroger to substantiate its advertisements (Tr. 2426). 224. The Burgoyne Checks could not have been relied upon as substantiation for Price Patrol advertisements (JX 1: Stec, p. 92) because the Burgoyne Checks were made approximately four times a year (Tr. 2112) while the Price Patrol advertisements were run weekly (Tr. 1653, 1844 2166-67 2439). Mr. Charles Thomas, Kroger Corporate Vice President, testified that the only weekly price check that was conducted by Kroger on a regular basis was the Price Patrol Survey (Tr. 2169). Mr. Thomas further testified that the only price check referenced in Price Patrol advertisements was the Price Patrol Survey, and consumers would have no way of knowing (62)anything about Kroger s Full Book or Burgoyne Checks (Tr. 2169-70). 225. Mr. Ted Hoover testified that 500 to 700 grocery items change in price each week (Tr. 1659), and Mr. Vaughn testified that prices of fresh meat change rapidly (Tr. 1947). Dr. Kohout stated that, in his opinion, the Burgoyne Check could not support a weekly claim (Tr. 1060-63). Mr. Hoover, testified that the Burgoyne Check was outdated by the time he saw the results (Tr. 1658). 226. Furthermore, in certain cities where the Price Patrol program was conducted, no Burgoyne Checks were made (Tr. 2429). For example, respondent's exhibit RX 978 shows that Burgoyne Checks were not made in cities such as Ada, Lawton, Duncan, Tyler THe, KROGER CO. b"" 639 Initial Decision Weatherford, Longview, Carrollon, and Albany (RX 978, pp. 1-2). These same cities ran the Price Patrol program (RX 900). CX 814 shows that only the Price Patrol, and no other check, was conducted in 43 cities in the Delta Marketing Area (Tr. 1332-33). Mr. Oliver Kroger s expert witness, confirmed that Burgoyne Checks were not conducted in 20% of the areas where the Price Patrol was run (Tr. 2786).

227. The Burgoyne Checks could not have been relied upon as substantiation for certain Price Patrol advertisements because, in certain cities where both checks were conducted, the competitors checked differed. For example, in Nashvile, Tennessee on June 22 1974, the Price Patrol checked the following competitors: Giant Foods, H. G. Hills, A&P, Food Town, Cooper Martin, and Big Star (RX 905). The closest Burgoyne Check (CX 282-22), June 27, 1974 checked the following competitors: Giant Foods, H. G. Hills, A&P Cooper Martin, Sunflower, and K-Mart. The Price Patrol showed Kroger lower than every competitor checked (RX 905), while one of the stores (K-Mart), omitted from the Price Patrol Survey, beat Kroger on the Burgoyne Check (CX 282-22). 228. In Indianapolis, Indiana on March 13, 1976, the Price Patrol checked the following competitors: Standard Del Farm, A&P, Thrif- Mart, Eisner, K-Mart, Marsh, and Preston Safeway (CX 125). The nearest dated Burgoyne Check, March 3, 1976 (CX 253-7), checked the following competitors: Marsh, A&P, Eisner, Standard, K-Mart and Thrif- T-Mart. The Price Patrol again showed Kroger lower than every competitor checked (CX 125), while the Burgoyne Check showed that Kroger was higher when compared to A&P and K-Mart (CX 253-7). In Bloomington, Indiana on January 31, 1976, the Price Patrol checked the following competitors: A&P, (63JStandard, Thrif- Mart, LG. , Marsh, and Eisner (CX 131) The nearest dated Burgoyne Check, January 16 1976 (CX 251--), checked the following competitors: Marsh, Thrif- Mart, Eisner, Standard, and A&P (CX 251--). Again the Price Patrol showed Kroger Jower than every competitor checked (CX 125), while the Burgoyne Check showed that Kroger was higher when compared to Standard (CX 251--). 229. The Burgoyne Checks could not have been relied upon as substantiation for certain Price Patrol advertisements because certain Burgoyne Check results were inconsistent with the results of the Price Patrol. Mr. Thomas testified that, from December 1972 until early 1978, each weekly Price Patrol Survey result in Atlanta showed Kroger winning the Price Patrol ('1r. 2140). Mr. Thomas also testified that, during the same time, certain Burgoyne Checks showed Kroger losing to one or more competitors on that check ('lr. Initial Decision 98 F. 2140-3). A review of Burgoyne Checks conducted in Atlanta during the period 1972 through 1978 shows numerous instances of Kroger losing to one or more competitors (CX 274-1 to 274-25; RX 925-1 to 925-3). Furthermore, Mr. Willie L. Greene testified that CX 813 showed numerous instances where the Burgoyne results were inconsistent with Price Patrol claims (Tr. 1342-55). 230. The annual cost to Kroger of conducting Burgoyne surveys weekly in 100 markets would be about $644 020 (RX 976; Hicks, Tr. 2337-38). This estimate is conservative because it excludes existing overhead costs such as Mr. Hicks' time, the time of other personnel who regularly evaluate the Burgoyne results, office space and similar administrative expense, as well as the substantial increase in working staff, management personnel and other resources which would be required for a massive expansion of the Burgoyne price survey program (Hicks, Tr. 2337; Benham, Tr. 2901-03). 231. It is found that with respect to comparative food price claims which are represented to be based on, or referring to, a survey, a reasonable basis means a reliable food price survey which is designed and conducted according to sound price survey procedures as recognized by experts who are trained and experienced in the science of survey methodology. The essential characteristics of sound food price survey procedures include aU the elements set forth in Part III A of the accompanying Order.

232. It is found that with respect to the advertising claims found to have been made by Kroger in the Summary Decision Order, dated May 17, 1978, Kroger did not possess (64Jand rely on a reasonable basis therefor. It is further determined that in the circumstances of this case those advertising claims were not only misleading and deceptive but also unfair to the consumer within the meaning of Section 5 of the Federal Trade Commission Act. Certain Practical Difficulties Attending Retail Food Price Surveys-FTC' s RFPS Experience A. Problems Related to Product Comparability 233. A fundamental problem in devising a survey to compare prices at competing food retailers is the variation in quality of items sold, especially in the case of fresh meat, produce and private label items (Oliver, 'Ir. 2576 in camera; 2643; RX 1003 in camera p. 45; Gast, Tr. 2209-13; Morgan, Tr. 696-708).

234. "Fresh meat" includes meat items which have neither been frozen nor processed by curing, smoking or other methods. The category includes items such as beef, pork, poultry, lamb, veal, and THE KROGER CO. 701 639 Initial Decision fish (Vaughn, Tr. 1886). It does not include processed products such as smoked and canned hams and bacon. Nor does it include frozen or canned meat (Vaughn, Tr. 1889-90).

235. The United States Department of Agriculture ("USDA") has promulgated standards for beef, veal and lamb. Those standards are in descending order of quality, USDA Prime, USDA Choice, USDA Good, USDA Standard, USDA Commercial, and USDA Canner and Cutter (Vaughn, Tr. 1898). USDA grades are made on the basis of marbling, conformation, age, and finish (Vaughn, Tr. 1902). 236. Less than 60% to 70% of beef slaughtered under federal inspection is graded (Vaughn, Tr. 1951). A large number of supermarket chains sell ungraded beef. These include A&P, The Eagle Division of Lucky Food Stores, and Chatham, a leading food chain in Detroit (Vaughn, Tr. 1901).

237. Kroger and many other chains follow a uniform industry nomenclature for most retail cuts of meat which they sell (Vaughn Tr. 1911-12). It is not, however, followed by all members of the industry (Vaughn, Tr. 1911-12). Not all stores that purchase graded meat disclose the grade at retail (Kohout, Tr. 1082; Vaughn, Tr. 1910). (65) 238. Kroger has a general policy of sellng USDA Choice beef (Vaughn, Tr. 1907, 1945). However, it buys only carcasses which would have qualified as "Choice" before the 1976 change; refuses to accept carcasses which fall in the low end of the Choice range (Vaughn, Tr. 1906-7). Kroger s purchasing specifications require that carcasses be in a certain weight range, exclude dairy type cattle and contain requirements on fat trim (Vaughn, Tr. 1907- 08).

239. Mr. Vaughn stated that he is not familar with the specifications or practices of other food retail firms for beef and could not say whether any of them may have additional requirements similar to Kroger s (Tr. 1907--8).

240. Variations in quality may arise from cutting methods and trim. For example, the way a Porterhouse or T-Bone steak is cut can affect value significantly. These cuts include a section of meat called the ((tail " which consists of rather stringy, tough meat surrounded by a great deal of fat. The "tail" can be trimmed off entirely or up to five inches of the tail can remain (Vaughn, Tr. 1915-19; Morgan, Tr. 69&-97).

241. "Trim" refers to fat left on a retail cut of meat (Vaughn, Tr. 1919). Kroger s specifications set a maximum of one-half inch external fat and one inch internal fat for beef; on pork the specifications set a maximum of one-quarter inch external fat Initial Decision 98 F. (Vaughn, Tr. 1920). Kroger takes a variety of steps to insure compliance with its trim specifications (Vaughn, Tr. 1920-21). some other chains have lower trim standards (Vaughn, Tr. 1921). 242. There are USDA quality grades for poultry, consisting of grades A, B, and C (with A being the top grade). These grades reflect the conformation and finish of the fresh bird and whether or not it is damaged (Vaughn, Tr. 1899).

243. Cutting methods on poultry can affect quality. For example, one cutting technique utilized by some of Kroger s competitors substantially lowers the ratio of usable meat to usable bone (Vaughn Tr. 1917 18).

244. The Department of Commerce has promulgated grades for seafood, but they are not widely used (Vaughn, Tr. 1899). 245. There are no USDA standards for pork (Vaughn, Tr. 1899 1934). Variations in quality in pork can arise from the age of the animal, as well as the cut and trim (Vaughn, Tr. 1916-17, 1920 1934-35). (66) 246. The phrase "produce" includes all fresh fruit and vegetables sold in the produce department of a grocery store (Bere, Tr. 1965). There are government standards for produce based on characteristics of the products and state of decay. These standards were set up many years ago to enable wholesale buyers and sellers to communicate over the telephone but have no meaning to consumers (Bere, Tr. 1965-9). Information on USDA grades for produce is often unavailable at retail (Bere, Tr. 1972-73, 1988-9 1991-92). 247. Kroger does not use the government grades for purchasing decisions, but relies instead on the reputation of the grower and the observations of experienced field buyers. Its general policy is to purchase the highest quality produce available (Bere, Tr. 1969, 1982- , 1994, 2000, 2025 2055; Morgan, Tr. 705). 248. The confusion inherent in the existing standards arises in large measure from the terminology used. For example, U.S. No. 1 is only the third highest grade for apples, but is the highest grade for peaches and onions; a U.S. Fancy apple is the second to the top grade while a U.S. Fancy potato is the top grade (Bere, Tr. 1967-68). 249. There are 6 USDA grades for apples, 8 to 10 different sizes and 8 to 10 varieties (e. Red Delicious) that appear in the market commonly. Quality distinctions are also drawn on the basis of origin. For example, a Western Red Delicious apple has different characteristics than an Eastern Red Delicious apple and commands a premium in the marketplace (Bere, Tr. 1970-71, 1974). Often the only information about apples available at point of retail sale wil be the price (which may be expressed per pound, per apple, or per , g., THE KROGER CO.

639 Initial Decision dozen), with no information about grade, size, variety, or origin (Bere, Tr. 1972-73).

250. Virtually all lettuce meets the U.S. No. 1 standard, which tells very little except the extent of decay and the presence of three wrapper" leaves on the head. Lettuce is usually sold by the head. A comparison of the price of two heads of U.s. No. 1 lettuce could involve heads which differ significantly in size, weight and quality (Bere, Tr. 1978-79, 1981, 1983--4).

251. Similar quality differences exist in the case of tomatoes (Bere, Tr. 1986-9), oranges and grapefruit (Bere, Tr. 1989-92), potatoes (Bere, Tr. 1992- , 2025- , 205&-7), strawberries (Bere Tr. 1999-2000), and (67Jpeaches (Bere, Tr. 1999-2001). Nearly all fresh fruits and vegetables in the produce department have quality differences of the sort described above (Bere, Tr. 1999). Even where products of substantially similar character are received by all competitors Chiquita bananas, the way in which the produce is handled can affect quality significantly (Bere, Tr. 1984, 2008-9 2048; Morgan, Tr. 697).

252. The Federal Trade Commission staff recognized the quality comparability problems inherent in a comparison of fresh meat and produce prices (RX 200 in camera pp. 6-7, and 11). 253. (See In Camera Findings) 254. The term "private label" may be defined as a line of products which are sold only by a particular company under various house brands. The products are manufactured by the food retailer procured from outside suppliers, or a combination of both (Gast, Tr. 2175- , 2178, 2181). Most large food retailing chains carry a private label line of products. Kroger s private label program is as large as or larger than most other food retailers (Gast, Tr. 2178-80). 255. The term "generic products" refers to products sold without brand name, private or national, which simply state Peas," on the label (Gast, Tr. 2181). Kroger does not have a generic program because it views generics as a merchandising gimmick (Gast, Tr. 2181--2). Several of Kroger s competitors carry generic products, including A&P, Weingarten, Liberal, Mad Butcher, Giant Eagle, and Topco Associates (Gast, Tr. 2183).

256. The United States Department of Agriculture has established grades for many food products canned vegetables, for which Kroger has private labels. The USDA grades for canned vegetables are, in descending order fancy, extra standard " and standard." (Gast, Tr. 2200-D1). The USDA grades are used primarily to enable retailers to communicate with suppliers concerning quality levels of products (Gast, Tr. 2200-2). The USDA grade is not Initial Decision 98 F. required to appear on private label products. The grades have not been used on product labels in part because the USDA nomenclature is not uniform and does not mean much to the consumer (Gast, Tr. 2203).

257. Kroger often carries two different private label brands for products which have varying quality levels. The "Kroger" brand is the top quality and "Avondale" is of lower (68jquality. Kroger buys fancy" grade vegetables for the "Kroger" label because its policy is to maintain a very high level of quality for its "Kroger" brand (Gast Tr. 2177, 2184, 2205-6, 2208).

258. Some other retailers may buy "fancy" for their private label line, but others may not. Kroger has no systematic way of determining what grades other retailers purchase. However, Kroger regularly checked private label products on both its Full Book and Burgoyne Price Checks and made judgments about its overall price position based upon these checks (Gast, Tr. 2216, 2208-9, 2223, 2347). 259. There are many products in Kroger s private label line for which there are no USDA grades. Kroger relies solely on its own specifications and quality control program for products without USDA grades (Gast, Tr. 2204-5).

260. For most of the products purchased from outside suppliers for private label programs, it is possible to buy goods varying widely in quality (Gast, Tr. 2184; Morgan, Tr. 708). Kroger has quality specifications which it provides to its suppliers for many private label products. These are designed so that quality wil be equivalent to the best-sellng "target" national brand (Gast, Tr. 2184, 2192- 2195- , 2206). Kroger has and enforces a quality assurance program designed in part to make certain that private label suppliers are meeting the Kroger specifications (Gast, Tr. 2184-6). It is not clear whether other food retailers have their own specifications for the private label products they purchase from suppliers (Gast, Tr. 2184-5).

261. In a typical metropolitan area in which Kroger operates there are likely to be at least three or four major competitors. Most if not all, of the competitors will have a private label program and many wil have two or three quality levels for some products. An individual with the experience of working in the private label program at Kroger for several years who went into a store and selected, for example, all the canned peas, private label and national brand, would be unable to make quality determinations from an examination of the labels (Gast, Tr. 2209-11). Price is not necessarily a good indicator of quality differences among private label products of competing food retailers (Gast, Tr. 2211-13). THE KROGER CO. 705 639 Initial Decision 262. Respondent's witnesses agreed that, in spite of these technical and sometimes significant difficulties (69)housewives face in comparing the quality of the products they purchase at food retail stores, housewives typically arrive at a pragmatic judgment regarding the product quality of a food retail store on the basis of their shopping experience over time (Tr. 1926, 2049- , 2060, 2213, 2223). Strict Random Sample Methodology Applied to Retail Food Price Survey May Be Difficult and Expensive 263. Since early 1973 the FTC staff has attempted to devise and recommend to the Commission appropriate methodologies for determining at which food store in a market the "consuming public. . . " (RX 200can satisfy their food needs at the lowest cost. . . . camera, p. 3). After years of substantial effort, the FTC staff recommendation is pending before the Commission (RX 1003 camera pp. 1--; Benham, Tr. 2936-38; Oliver, Tr. 2575 in camera, 2649, 2707).

264. In 1972 the FTC staff investigated low-price claims advertised by food retailers, including the claim "Kroger Is The Low Price Leader" made in Dallas (RX 1003 in camera p. 21). In January 1973 the FTC announced publicly a "Proposed Protocol For Federal Trade Commission Retail Food Price Survey" (hereafter "RFPS") (RX 200 in camera). The FTC stated that it would conduct its own retail food price surveys in several markets, comparing the price levels among stores in each market. The results of the surveys were to he used in enforcement actions against " significant false advertisers " and also published for the benefit of consumers (RX 200 in comera p. 3). The FTC anticipated that the "first survey wil be initiated sometime in 1973. " (RX 200 in camera p. 6).

265. From their inception, the FTC surveys were to employ random sample methodology and probabilstic techniques (RX 200 camera, pp. 6-). Two alternative methodologies were described in the FTC January 1973 announcement, Method A and Method B (RX 200 in camera pp. 6-8). Only Method B was expected to provide " statistically valid sample of the items likely to be purchased by a typical consumer" and therefore only the market basket totals from Method B would be "indicative of the overall price levels of each store. " (RX 200 in camera pp. 11-12). 266. (See In Camera Findings) (70) 267. (See In Camera Findings) 268. (See In Camera Findings) 269. (See In Camera Findings) 270. (See In Camera Findings) Initial. Decision 98 F. 271. (See In Camera FindingsJ 272. (See In Camera FindingsJ 273. (See In Camera FindingsJ 274. (See In Camera Findings J 275. (See In Camera Findings) 276. (See In Camera FindingsJ 277. (See In Camera FindingsJ 278. (See In Camera FindingsJ 279. (See In Camera Findings) 280. The record does not show whether or not any other entity has conducted retail food price surveys to measure price levels at competing retailers by using random sample methodology (RX 992 , No. 73; RX 1003 in camera pp. 25-26; RX 207 in camera pp. 9 10).

281. One of the theoretical flaws of the RFPS methodology is that a food retailer can improve its standing against competitors in the survey by carrying low quality items at a lower price, whether there were any sales of these items or not (RX 1003 in camera pp. 25-26; RX 207 in camera pp. 9-10).

282. Another flaw is that a food retailer could improve its standing . against competitors in the . survey by carrying very. large sizes of items at relatively low average prices per pound, whether or not there are any sales of such items to consumers (Oliver, Tr. 2781). 283. Another flaw is that RFPS results would be affected solely by differing product mixes. For example, a store which carries domestic brands of tea at a low price (71Jand a wide variety of more expensive foreign teas would be rated as more expensive than a store which carries only domestic tea at a price higher than the first store (RX 1003 in camera pp. 3 0; RX 207 in camera, pp. 51-52). 284. (See In Camera FindingsJ 285. Kroger ran a survey in St. Louis to test the practicability of a random-sample, product type survey of the sort envisioned by both the RFPS methodology and Dr. Kohout's testimony ("St. Louis Survey ). This survey showed that the fears about the effect of product mix on RFPS results were well-founded (Oliver, Tr. 2600-1 in camera). In the St. Louis Survey 12 products, analogous to the products" of the RFPS, were selected (RXs 981-985). The selection of products in the St. Louis Survey was conducted by random sampling, except for the meat product (Oliver, Tr. 2594-2600 camera). Prices for all items constituting each of these 12 products were surveyed at two stores of each of five competitors, including Kroger (RXs 981-985).

286. When average price per pound for a product was computed THE KROGICR CO. 707 639 Initial Decision for each store in the St. Louis Survey, the relative position of one store of the same chain was different from that of the chain s second store, even though all items common to the two stores were priced the same. These differences occurred solely because of differences in product mix between the two stores of the same chain (Oliver, Tr. 2601-D7 in camera).

287. The random sample methodology recommended by Dr. Kohout for conducting retail food price surveys is in all essential respects the same as the RFPS methodology (Kohout, Tr. 1053, 1068- , 1075, 1078, 1169- , 1178, 1195- , 1207, 1231, 1237- , 1270 1279-80; Oliver, Tr. 2567 , 2593-94 in camera 2645; RX 1022 camera; Benham, Tr. 2934). Dr. Kohout reviewed a description of the survey design employed by the RFPS, and agreed it accorded with his description of an appropriate random sample methodology for conducting retail food price surveys (Kohout, Tr. 1068-70, 1244-7; RX 1022 in camera).

288. Dr. Kohout testified that the methodological difficulties presented by quality differences in items between stores should be ignored (Kohout, Tr. 1074, 1262; RPF 236 in camera). The apparent basis for this position is his assumption that, when reviewing survey results, consumers would simply assume that we are dealing with (72Jprices and not with quality. " (Kohout, Tr. 1321). Another rationale proposed by Dr. Kohout for ignoring quality was to assume that quality differences in items occur randomly across items and among stores without any correlation with price (Kohout, Tr. 131&- , 1322-23).

289. According to Dr. Kohout, in a retail food price survey using random sample methodology with a population of products rather than items, a price per ounce would be computed for each store for each product from the surveyed price data for items in each product (Kohout, Tr. 1075, 1197, 1207, 1212-13). According to Dr. Kohout there are three !!equally acceptable" methods for computing price per ounce for each product: (1) using the lowest price per ounce for any item found for a specific product; (2) computing the average price per ounce for all items found for a specific product; and (3) for any product, discarding the item with highest price per ounce discarding the item with lowest price per ounce, and then computing the average price per ounce for all remaining items found for a specific product (Kohout, Tr. 1204-5, 1213-14). 290. The St. Louis Survey showed that the results of a retail food price survey employing random sample methodology with a population of products rather than items can vary substantially depending on which of Dr. Kohout's three "equally acceptable" methods for Initial Decision 98 F. computing price per ounce is employed (Oliver, Tr. 2603-07 camera; 2778-79; RX 985). This methodological problem is inherent in any product-type retail food price survey employing random sample methodology (Oliver, Tr. 2606-7 in camera). 291. Dr. Kohout's evaluation of the methodology of various types of retail food price surveys, including the Price Patrol Survey, the Burgoyne survey, the full book check and the RFPS, rests on his conclusion that the 'tfar superior" methodology for conducting retail food price surveys is random sample methodology (Kohout, Tr. 1047 1054, 1056, 1060-2, 1065, 1075- , 1083- , 1091- , 1109- 1149- 1281 1301-08).

292. Dr. Kohout had no knowledge of the Burgoyne surveyor full book check methodologies, nor any knowledge of the manner in which Kroger evaluates the results of these surveys (Kohout, Tr. 1295-96).

293. Dr. Kohout was not provided any information about the RFPS experience by complaint counsel, except for the information contained in a description of the RFPS survey design (Kohout, Tr. 1164 , 1246; RX 1022 in camera). In testifying that a retail food price survey employing the (73)essential aspects of the RFPS methodology would be feasible and inexpensive to accomplish (Kohout, Tr. 1064 1305-08), Dr. Kohout was therefore unaware of the experience of the FTC with the RFPS.

294. (See In Camera Findings) 295. The annual cost of conducting the food component of the Consumer Price Index-a retail food price survey of far less complexity than the RFPS and which does not comply with the methodological standards of Dr. Kohout and the RFPS-n a monthly basis in 56 cities with a six-week turnaround time for computation of results was estimated as $1.8 millon by Dr. Benham (Tr. 2900, 2905, 2921, 2942).

296. (See In Camera Findings) 297. Since the RFPS methodology results in unique market basket totals for every store in the market, even where a chain has the same price for all items carried in any of its stores in the market at least several stores of a chain in a market would have to be surveyed as part of an RFPS survey (RPF 247; Oliver, Tr. 2579-80 2605 in camera). Since the ongoing costs of RFPS surveys are directly related to the number of stores included in each survey, the annual cost estimates of conducting RFPS surveys discussed above, which assume only a single store of each chain need be surveyed, are conservative (RX 240 in camera p. 3; RPF 259-262; Oliver, Tr. 2580 2605 in camera).

THE KROGER CO. 709 639 Initial Decision 298. (See In Camera FindingsJ 299. (See In Camera FindingsJ 300. (See In Camera Findings) 301. (See In Camera Findings) 302. (See In Camera Findings) VI. Economic Effects Of Restraints On Retail Food Price Advertising 303. Advertising which is not deceptive or misleading and which provides consumers with useful information concerning product characteristics and available price alternatives (74Jreduces the search time and search costs which consumers otherwise would have to incur to gather that information (Benham, Tr. 2821- , 2855- 3081-84).

304. Dr. Benham, Kroger s expert witness, testified that comparative price advertising can provide consumers with useful information even if that information is not "perfectly accurate" in the sense of being based on surveys conforming to strict theoretical standards of statistical methodology, provided its limitations and "imperfections" are clearly stated in such advertising (Benham, Tr. 2858, 3071 3080-84).

305. The availability to consumers through truthful or nondeeeptive advertising of more price information at less search cost is thought to make the marketplace more competitive and lead to lower prices (Benham, Tr. 2821- , 2828-29, 2831- , 2838-39, 2844- 2855--7 2977--0; RX 207 in camera p. 9). 306. According to Dr. Benham, the true total cost of purchasing a product includes both the direct purchase price of the product, and the time and cost of learning about the characteristics and price of the product (Benham, Tr. 2821, 2856-57). Therefore, to the extent the availability of price information through advertising is reduced the true total purchase costs of goods to consumers are likely to rise because both the prices of the goods and the search costs for obtaining information will increase (Benham, Tr. 2821- , 2856-57). 307. Increased availability of information through advertising makes it more likely that price-competitive firms will succeed in the marketplace and less likely that higher priced firms wil maintain or increase their share of the market (Benham, Tr. 2822, 2829- , 2846 2969- 2975-76).

308. Increased restraints or costs imposed on price advertising may make it more difficult for highly price-competitive and more efficient firms to convey their pricing posture to consumers, thus 710 FEDERAL 'fHADE COMMISSION DECISIONS Initial Decisioo 98 F. discouraging such firms from entering new markets (Benham, Tr. 2822, 2829- , 2832- , 2846, 2866, 289&-99, 2941, 2964, 2969- 2973-76).

309. According to Dr. Benham, to the extent that imposition of quality standards on advertising of price information through substantiation and disclosure (75)requirements increases costs, it may reduce the amount of such advertising and thus reduce the amount of price information available to consumers (Benham, Tr. 2819 2824 2863--4 2867 3073-75).

310. The cost estimates for the FTC retail food price survey, the pilot St. Louis Survey conducted by Kroger following Dr. Kohout's proposed methodology, and the government' s CPI experience all tend to show that it would cost a large amount of money to set up and run such surveys in the number of cities where Kroger ran the Price Patrol (Benham, Tr. 2900-01).

311. According to Dr. Benham, the costs of a survey meeting the setrequirements of the definition of "methodological soundness" forth in RX 1007, part I C, would be exponentially greater than the cost of Price Patrol (Benham, Tr. 2899-2900, 3031-33). 312. Respondent's employees and its expert witness testified that, because of the expense involved, food retail firms would not attempt to conduct retail food price surveys and advertise their results if the surveys were required to comply with a strict definition of methodological soundness such as that set forth in part I C of RX 1007 (Benham, Tr. 2897-99; Thomas, Tr. 2115-16; Stec, Tr. 177&-79). 313. Dr. Benham also testified that the result of requiring any published retail food price survey to meet a strict definition statistical "methodological soundness" wil be to reduce the amount of useful information provided to consumers, which in turn would result in higher prices to consumers in general, a competitive disadvantage for the most price-competitive firms, and a relatively greater adverse cost impact for the least sophisticated consumers (Benham, Tr. 2863--4, 2867, 2898- , 2969-72), DISCUSSION A. A Reasonable Basis For A Comparative Food Price Claim Purportedly Based On Or Referring To A Price Survey The Commission s requirement that an advertiser must possess and rely on a reasonable basis for an affirmative product claim has been judicially sanctioned. Pfizer, Inc" 81 F. C. 23 (1972); Firestone Tire Rubber Co. 81 F. C. 398 (1972), aff'd 481 F.2d 246 (6th Cir. 1973), cert. denied 414 U. S. 1112 (1973); National Dynamics Corp., 82 , THE K.KUli.rn L.V.

639 Initial Decision C. 488 (1973), aff'd 492 F. 2d 1333 (2d Cir. 1974), cert. denied, 419 U.s. 993 (1974). (76) The basic rationale of Pfizer is that an affirmative product claim carries with it an implied representation that the advertiser possessed and relied on a reasonable basis for the claim when the claim was made and that such an advertising claim in the absence of a reasonable basis is an unfair act or practice in violation of Section 5 within the meaning of Section 5. See FTC v. Sperry Hutchinson Co. 405 U.s. 233, 234 (1972). The reasonable basis requirement applies even if an advertisement claim is in fact true. 81 F. C. at 63. Also see id. at 67-68.

In determining what constitutes "a reasonable basis " the Commission set forth a number of guidelines in Pfizer. First, the Commission made it clear that the requirement is not solely a "reasonable man test. The reasonable basis requirement questions both the reasonableness of an advertiser s actions and the adequacy of evidence upon which such action is based. The reasonable basis standard is essentially a fact issue to be determined on a case by-case basis, and depends on such overlapping considerations as: (1) the type and specificity of the claim made; (2) the type of product; (3) the possible consequences of a false claim; (4) the degree of reliance on the claim by consumers; and (5) the type and accessibility of evidence adequate to form a reasonable basis for the particular claim. For some types of claims and for some types of products, the only reasonable basis " fairness and in the expectation of the consumers" would be an adequate and well-controlled scientific test. 81 F. C. at 64, 66-67. This proceeding involves comparative food price claims for specific markets which are represented to be proven by price surveys. In my view, in the circumstances of this case, such comparative price claims constitute in fairness and in the expectation of consumers and as a matter of law, an implied representation that the advertiser possessed and relied on a reliable food price survey which in fact proves those claims. Cf Standard Oil Co. of California 84 F. 1401 1472 (1974), modified on other grounds 577 F.2d 653 (9th Cir. 1978). The record is clear that in order for a food price survey to prove a comparative price proposition, the survey must be based on sound price survey procedures as recognized by experts trained and experienced in the science of survey methodology. The record reflects a reasonable agreement among such experts as to the essential elements of a reliable food price survey (F. 139-144, 178 198). And there is little doubt that such a survey can be done although it may be expensive and difficult. (77) 712 FEm;RAL TRAm; COMMISSION DECISIONS Initial Decision 98 F. B. Relief Respondent' s acts and practices found to be in violation of Section 5 of the Federal Trade Commission Act fall into three general groups: (1) express and implied advertising representations that the Price Patrol Survey proved certain comparative price claims in certain cities or metropolitan areas, (2) express and implied advertising representations that the Price Patrol Survey means shopping at Kroger wil result in overall savings, and (3) implied representations that Kroger possessed reasonable bases for these comparative price the relief must ban all comparative price claims. Therefore first claims not based on a reasonable basis, regardless of whether a particular claim is purportedly based on a price surveyor not. Part II of the Order is designed to accomplish that objective. Kroger argues that the lack of a reasonable basis was not specifically alleged with respect to lower price claims which are not purportedly based on any price survey and that, therefore, imposition of a reasonable basis requirement for all comparative price claims is overbroad and unreasonable. However" Kroger s advertising claim that the Price Patrol proved Kroger had lower prices or was a low price leader in a city necessarily implied a broad claim of lower prices apart from the Price Patrol Survey. Under these circumstances, the reasonable basis requirement for all types of comparative price claims is clearly and reasonably related to the violation found and is eminently justified. Alternatively, an acrossthe-board reasonable basis requirement for any comparative food price claim for a city is necessary and appropriate as a " fencingprovision in view of the specific violation which was alleged and proved with respect to comparative price claims purportedly based on surveys. Jay Norris Corp. v. FTC, 598 F.2d 1244 (2d Cir. 1979) (2d Cir. No. 78--151, decided May 1, 1979).

Secondly, with respect to any comparative price claim for a city which is purportedly based on a price survey, Kroger must possess and rely on a survey which has all the essential characteristics of sound price survey procedures as generally recognized and understood by the community of experts trained and experienced in survey methodology. Otherwise, any express or implied claim that a price survey "proves" or "shows" anything about comparative prices in a given city would be patently misleading, deceptive and unfair. Part III A of the Order is designed to accomplish that objective. The record is clear that a competent and (78Jreliable food price survey which meets the conventional technical requirements regarding survey design and execution can be done, although such a survey THt. h..tul .tn \A...

639 Initial Decision may require substantial expenses and encounter certain practical diffculties (F. 139, 233-297). And, there is no cogent reason why Kroger should be allowed to make a comparative food price claim which is represented to be based on a food price survey without in fact having and relying on a competent and reliable survey, which meets these technical requirements. If Kroger is not willing to expend the necessary time and financial resources, it should simply forego comparative price claims represented to be based on, or referring to, a food price survey.

Kroger s argument that such a requirement would exacerbate the inflation in food prices as a result of food retailers' passing on the mandatory survey expenses to the consumer or by inhibiting comparative price advertising by food retailers is unpersuasive. Needless to say, inflation is the most important and urgent domestic issue the Nation faces today and the Federal Trade Commission should not do, in the guise of trade regulation, anything that may compound that problem.' However, unless Section 5 of the Federal Trade Commission Act is to be set aside for the duration, a clear statutory violation affecting the consumer in such an important area as food purchases must be stopped, especially where, as here, the affected advertising claim (that a price survey proves the advertiser has lower prices in a market) is by no means the only, or the customary, way of comparative food price advertising. The record as a whole does not provide a solid basis for concluding that the requirements set forth in III A wil operate to inhibit all or most comparative food price advertising or that a large number of food retail firms will insist on our survey-proves type advertising and pass on the incremental survey expenses to the consumer. The record does show, however, that Kroger s Price Patrol advertising program was very effective and instrumental in markedly increasing Kroger s market shares (F. 130 supra). In these circumstances Kroger s argument must be rejected as essentially self-serving and unpersuasive. (79) Thirdly, on the other hand, Kroger should be free to make comparative food price claims for any city as long as (1) it does not say that these claims are proved by or based on a price survey and (2) it clearly and fairly discloses the limitations of the comparative claims resulting from any short cut employed in the survey procedure. Part III B (1) and (2) of the Order are designed to , Mr- R. Robert RusscJJ, Deputy Director, Counci! on Wag and Price Stability, in a letter dated December I 1978 and addressed to the Commission Secretary, expressed, without attempting to "judge the merits of this proceeding," the Council's concern t.hat a requirement for more rigorous and extensive survey procedures may inhibit dissemination of price information, discourage pric compttiLion and lead to higher foo prices. These views were taken intocoflsideration by me.

367- 44'1 0 - 82 - 45 QL 3 Initial Dccision 98 F. accomplish this objective. The record clearly shows that comparative price information enhances competition, provides useful information to the consumer and otherwise inures to the benefit of the consumer (F. 303-308 supra). The record also indicates that a technically competent and reliable price survey may be expensive and difficult fraught with many problems peculiar to the food retail business. In these circumstances, the question whether food retailers should be allowed to make any comparative price claims at all without a competent and reliable price survey is a policy question for the Commission to decide. In my view, common sense would dictate that comparative price advertising of the type under discussion here should be allowed in the absence of a technically competent and reliable survey so long as such advertising alerts consumers to the limitations of the claim. Although this approach is not without certain problems, including some practical and technical problems related to the disclosure requirements ' I am persuaded that, on balance, it is a reasonable and realistic way of allowing comparative food price advertising that is not misleading, deceptive or unfair. Kroger s argument that any disclosure requirement will, as a practical matter, discourage all comparative food price advertising is not persuasive. In any event, a specific comparative food price advertising in the absence of any price survey would be contrary to the Commission s established policy of requiring reasonable substantiation for advertising claims. The record does not show any reason why the retail food industry should be an exception to that policy. In my view, a reasonable substantiation with respect to specific comparative price claims (not purportedly based on a price survey) in the food retail business means some kind of bona fide price survey, accompanied by a fair and straightforward statement of the survey limitations. (80) Complaint counsel have proposed alternative requirements with respect to comparative price claims not represented to be based on or referring to, a price survey. They include the use of "market basket surveys " and number-of-item-surveys.)) See I and IV of Complaint Counsel's Proposed Order. They appear to be plausible and may be reasonable and appropriate for the purposes of this case. However, this record does not provide a sufficient basis for an informed and reasoned judgment regarding these alternatives. On the basis of this record, I am unable to evaluate the relative merits and shortcomings of the proposed or other alternatives or assess 1 For example, I am aware t.hat the di"dosure requirements may raise substantial technical problems for radio and television commercials depending on t.he deg-ree of imperfections present in a particular survcy. It may be that the electrunic media arc not the most idea! media for 3 broad comparative foo price claim of the sort under discu5Siofl.

THE KROGER CO.

639 Initial Decision their economic and marketing ramifications with any reasonable degree of confidence. For these reasons, I am unable to accept or endorse complaint counsel's proposed alternatives. Fourthly, Kroger should not be allowed to make comparative food price claims with regard to a number of items or groups of items (e. the Price Patrol advertisements) without stating that such claims do not mean that the consumer will always save by shopping at Kroger for the simple reason that such a claim necessarily implies a claim that consumers will save overall by shopping at Kroger. Without a straightforward disclaimer statement, the claim will be misleading and deceptive. III B(3) of the Order is designed to accomplish that objective. It is of course arguable that many housewives who view such advertisements wil know, or should know, that whether they will save at all, or how much they will save, by shopping at Kroger will depend on what they buy at Kroger. However caviat emptor inconsistent with the basic purposes of Section 5. Indeed, Section 5 was meant to protect not only the well tutored and intelligent but also the ignorant, the gullble and the credulous. FTC v. Sterling Drug, Inc. 317 F.2d 669 , 674 (2d Cir. 1963); FTC v. Standard Education Society, 302 U.S. 112, 116 (1937). Respondent vigorously argues, on the basis of the testimony of Dr. Benham, its expert witness, and certain corporate employees, that any requirement for disclosure of survey limitations in the same advertisement will, as a practical matter, inhibit all comparative price advertisements to the detriment of competition and consumer interests. However, Dr. Benham also stated that he did not mean to advocate or condone deceptive or untrutbful advertising (Tr. 3081- 84). And the record clearly shows that comparative price claims not accompanied by appropriate disclosure of limitations are misleading, deceptive and unfair. This fact cannot be ignored or wished away. Comparative price (SlJadvertising is desirable only when it is not misleading, deceptive or unfair.

It should be pointed out here that the Order leaves Kroger free to make any general and non-specific comparative price claim as long as it is not related to a particular city, metropolitan area or a competitor(s). With respect to such general claims, no price survey of any kind, and no disclosure statement, would be required. Whatever a reasonable substantiation may be for such general claims, this record convinces me that such substantiation need not include a price survey. The record also shows that food price surveys are more complex and diffcult than meet the eye and that it should not be required lightly of food retailers. In this connection, complaint counsel's proposed disclosure requirement (V B of complaint coun- Initial Decision 98 F.T. sel's Proposed Order) is rejected. Such a requirement is inconsistent with the determination that reasonable substantiation for general claims involved here need not include a price survey. In any event, I am unable to conclude on the basis of this record that such a general claim will be misleading, deceptive or unfair unless accompanied by the proposed disclaimer or that the absence of a price survey in these circumstances is a material fact which must be disclosed. Kroger s constitutional objections to the reasonable basis requirement are without merit. The reasonable basis requirement with respect to affirmative product claims has been judicially sanctioned. And in appropriate cases, a reasonable basis requires a competent and reliable scientific test. Firestone Tire Rubber Co. v. FTC, 481 2d 246 (6th Cir. 1973), ccrt. denied 414 U. S. 1H2 (1973); National Dynamics Corp. v. FTC, 492 F. 2d 1333 (2d Cir. 1974), ccrt. denied 419 U.s. 993 (1974). With respect to comparative food price claims, a reaonsable basis would be a competent and reliable food price survey, or a bona fide food price survey accompanied by complete and fair statement of the survey s limitations where short cuts are material limitationsemployed, so as to alert the consumer to any the conclusions that can be validly drawn from the survey. Standard , 1472 (1974), modified on otherOil Co. of California 84 F. C. 1402 grounds 577 F. 2d 653 (9th Cir. 1978). Cf Bristol-Myers Co. 46 F. Inc. 162 (1949), a/i'd 185 F. 2d 58 (4th Cir. 1950); Country Tweed, FTC, 185 F. 2d 144 (2d Cir. 1964); General Motors Corp. 66 F. C. 267 272 (1964). Kroger has not pointed to, and I am not aware of, any case where a court has held that the First Amendment insulates from prior restraint misleading, deceptive or unfair advertising claims violative of (82)Section 5 of the Federal Trade Commission Act. And comparative price claims involved in this proceeding would be clearly misleading, deceptive and unfair unless the advertiser has a reasonable basis which meets the requirements set forth hereinabove.

Kroger s argument that the Commission s prosecution of this case constitutes an unlawful abuse of discretion is without merit. Although the Commission could have proceeded by way of rulemaking to deal with the matter of comparative food price advertising or awaited the conclusion of the pending Proposed Protocol For Retail s discretion to prosecuteFood Price Survey (HFPS), the Commission the instant litigation is well established. Moog Industries v. FTC, 355 where the S. 411 (1958). This is not Marco situation, Commission sought to prohibit practices not shown to be misleading . Man:oSale. Go. v. FT, 453 F.2d 1 (2dCir. 1971J THE KROGER CO. 717 639 Initial Decision or deceptive and do so without adequate articulation of its reasons when a Trade Regulation Rule adopted contemporaneously by the Commission did not ban similar practices. Finally, Kroger s argument based on discontinuance is rejected. The record is clear that the Price Patrol advertisements were highly effective and led to marked increases in Kroger s market shares in many markets (F. 130, 135). The record is also clear that the primary reason for discontinuing the Price Patrol program was the pendency of this litigation (F. 107). Under these circumstances, Kroger discontinuance argument is entirely unpersuasive. CONCLUSIONS OF LAW 1. The Federal Trade Commission has jurisdiction over respondent' s advertising program, known as the Kroger Price Patrol, under Section 5 of the Federal Trade Commission Act. 2. The Complaint herein states a cause of action, and this proceeding is in the public interest.

3. Respondent's use of deceptive and unfair advertising representations has had, and now has, the tendency and capacity to mislead and deceive a substantial portion of the purchasing public into the mistaken belief that said representations are true, and to lead them into the purchase of substantial quantities of respondent's goods by reason of said mistaken belief. (83) 4. The acts and practices of respondent, as found herein, were and are prejudicial and injurious to the public, and constitute unfair or deceptive acts or practices in or affecting commerce, in violation of Section 5 of the Federal Trade Commission Act. 5. The accompanying Order is necessary and proper for the purpose of prohibiting the continuation of the proscribed acts and practices and is in the public interest.

ORDER Definitions A. Respondent means The Kroger Company, a corporation, its successors and assigns, and its officers, representatives, agents and employees, acting directly or indirectly through any corporation subsidiary or other device in the sale of food, household items and other merchandise in or affecting commerce, as "commerce" is defined in the Federal Trade Commission Act. Initial Decision 98 F. B. Comparative retail food price claim means any advertising claim which represents, directly or indirectly, that Respondent' retail food stores have lower or lowest prices, or more lower or lowest prices, either overall or with respect to any item, groups of items or product categories.

II.

It is ordered That Respondent cease and desist from advertising, directly Of indirectly, any comparative retail food price claim which refers, directly or indirectly, to a (84)particular city, metropolitan area or a competitor (or competitors), by name or other designation unless:

A. Respondent possesses and relies on a reasonable basis therefor at the time such claim is made;

B. Any results of the surveyor study which are advertised, and any claims based on such surveys or studies, as presented in a manner that fairly and impartially represents those conclusions that may be validly drawn; and C. The results, data and complete description of the method and procedures used in the price survey, study or check related thereto are made conveniently available to the public for inspection and copying at Respondent' s food retail stores. III.

For the purposes of this Order A. With respect to any comparative retail food price claim which refers to, or is represented to be based on, any survey, study or check a reasonable basis shall mean a retail food price survey which is designed and conducted according to sound survey procedures. Sound survey procedures" shall include the following characteristics: (85) (1) Selection of Hems.

(a) Surveyed items or products must be selected in such a way as to provide a valid basis for generalizations to be made from the survey results regarding the prices of products customarily sold at a retail food store. Items must be selected from all major product categories to which the survey results are generalized. (b) If the items are selected by Respondent, they must be selected randomly from the total population or randomly from each major Tm; KROGER CO. 719 639 Initial Decision category of products. Respondent shall not make the list of items available to persons responsible for pricing decisions before the survey is made. Alternatively, the items may bc selected by an independent outside firm. either randomly or in such a way as to make the items fairly representative of the total population. If the items are selected by an independent outside firm, the survey item list shall not be disclosed to any firm to be surveyed before the survey is conducted. Rer:ardless of whether items are selected by Respondent or by an independent outside firm, the survey item lists must be periodically rotated. (86) (2) Comparability of Survey Items.

Items which are compared must be identical or substantially similar. Branded items which are the lowest priced items available in a product category are substantially similar. Meats identical in cut and grade are substantially similar. Produce items of the same type and geographic origin are substantially similar. If an identical or substantially similar item is unavailable at a particular store, the item must be excluded from the survey for all stores and shown as unavailable for comparison.

(3) Stores Surveyed.

Only one store of each firm need be surveyed in a single survey period in a single city or metropolitan area. If a firm has price zones among its stores in the same city or metropolitan area, the store of that firm should be periodically rotated among price zones. (4) Competitors Surveyed.

Competitors to be surveyed must be selected in such a way as to provide a valid basis for any generalization to be made from the survey results. (87) (5) Reasonable efforts must be made to avoid conducting surveys under circumstances which may introduce systematic error in survey results.

B. With respect to any comparative retail food price claim, made directly or indirectly, which refers, directly or indirectly, to a particular city, metropolitan area or competitor(s) and which does not refer to or is represented to be based on any survey, a reasonable basis shall mean:

(1) A price survey which is designed and conducted according to sound survey procedures as set forth in III A (1) throur:h A (5) hereinabove, or (2) A price survey which does not rneet one or more requirements Initial Decision 98 F. set forth in III A (1) through A (5) hereinabove provided, however complete and fair disclosure of the limitations due to any failure to meet such requirements is made in the same advertisement. For example, if the survey excluded any major product category, such as meat, the advertisement must contain the following disclosure: THE SURVEY ON WIICH THIS ADVERTISEMENT IS BASED DID NOT IN- CLUDE MEAT lCATEGORY(IES)j.

(3) With respect to a comparative food price claim for a group of items which, directly or (88Jindirectly, purports to show which competitor offers more lower prices, the advertisement must contain the following disclosure:

WHETHER YOU WILL SAVE OR HOW MUCH YOU WILL SAVE WILL DEPEND ON WHAT YOU BUY.

IV.

Disclosures required in this Order shall be featured prominently (A) in all printed advertisements and shall be printed in at least sixteen (16) point, bold-face capitals, (B) in the audio portion of any radio advertisement and no other sounds, including music, shall occur at the same time, and (C) in the visual portion of any television advertisement throughout the length of the commercial. It L, further ordered That Respondent shall, within sixty (60) days after service of this Order upon it, file with the Commission a written report setting forth in detail the manner and form in which it has complied, or intends to comply, with this Order. VI.

It is further ordered That Respondent notify the Commission at least thirty (30) days prior to any proposed change in the corporation such as dissolution, assignment, or sale, resulting in the emergence of a successor corporation, the creation or dissolution of subsidiades or any other change in the (89Jcorporation which may affect compliance obligations arising out of this Order. VII.

It is further ordered That the allegations contained in Paragraphs .L.L.L'- .".,..U.LJ., VV. 1L..1 639 Opinion Six D and Seven D of the Complaint be, and they hereby are dismissed.

OPINION OF THE COMMISSION By CLANTON Commissioner:

The Kroger Company, a retail food chain, is charged with engaging in unfair and deceptive practices as a result of its alleged failure to substantiate certain claims made in a comparative price advertising campaign kllOwn as the Kroger Price Patrol. During this campaign Kroger organized small teams of shoppers into Price Patrols, which conducted weekly surveys of the prices charged for specific products by Kroger stores and their local competitors. The company advertised the survey results, which in the vast majority of occasions showed Kroger with lower prices on substantially more items than its rivals.

On complaint counsel's motion for summary judgment, the administrative law judge ("AI..") found that the advertisements had represented, inter alia, that the Price Patrol surveys were methodologically sound and that they proved both that most items in Kroger stores were cheaper than in competitors' stores and that consumers would save overall by shopping at Kroger. Evidentiary hearings were then held with respect to the remaining issues, such as the adequacy of respondent's substantiation, the adequacy of its disclosure of material facts, and the nature of appropriate relief in the event respondent was found to have violated the FTC Act. (2)In his Initial Decision, issued on June 11, 1979,' the ALJ determined that the respondent' s substantiation for its Price Patrol claims was deficient. He concluded the company had violated Section 5 of the Federal Trade Commission Act and ordered it to desist from further comparative food price advertising unless it followed specified survey procedures and made disclosures qualifying its claims. L The Initial Decision ar,d other materials ;n the record of this case will be cited to with the followin.: abbreviations Initial Decision, l"finding No ID p. Initial Decision, Page No Complaint ('.."onsel Exhibit N"

Respondenu;' t:xhibit No T,. Transcript of Testimony, Page No. RAR Respondents' Appeal Brief, Page No CAR Complaint Counsel's Answering Brief, Page No. RRB Respondents' Reply Brief, Pagl! No CPF Complaint Counsel's Proposed Finding, Page No. RPF Respondents' Proposed Finding, Page No TROA Transcript of Oral Argument, Page No Opinion 98 F.

Kroger has appealed, attacking the proceedings below on a variety of fronts. It argues first that the ALJ's interpretation of the advertising claims was incorrect and that his decision to interpret the advertisements on a motion for summary judgment was erroneous. Second, Kroger contends that even if its advertisements made the representations found by the ALJ, they were reasonably substantiated by the Price Patrol surveys themselves and by other Kroger-sponsored comparative price checks. Third, the respondent objects to the ALJ's order, asserting that its stringent survey and disclosure requirements will only discourage Kroger and its competitors from running any future comparative food price advertisements thus depriving consumers of useful information and weakening competitive pressures to reduce food prices. Finally, Kroger raises several evidentiary issues and argues that ALJ Hyun should have been disqualified from trying this case.

These contentions will be addressed in turn, following an outline of Kroger s pricing policies and its Price Patrol advertising campaign. BACKGROUND Kroger is one of the largest retail food companies in the nation. At the time this case was tried, Kroger was doing business in twenty states, operating 1170 stores which it had grouped for organizational purposes into thirteen Kroger Marketing Areas ("KMA"). (3) Prior to 1972, Kroger s prices on most shelf goods were relatively high, although it featured certain highly discounted, or "deep cut products in its advertising to attract customers. (Tr. 2315.) In May 1972, however, the company decided to switch to what it calls its Everyday Low Price policy ("EDLP"), under which all groceries and merchandise were to be sold at reduced prices. The objective of the Everyday Low Price policy was to ensure that a person who shopped at Kroger would, over time, realize savings that matched or surpassed those offered by comparable stores. (Tr. 2313-15. Thus, Kroger s Grocery Merchandisers, who price the goods to be sold in their respective KMAs were instructed to reduce shelf prices to those charged by the lowest merchant in town that offered the same services. (CX 15 p. ) The Everyday Low Price policy relied heavily on merchandisers' use of rollers-these are typically highvolume, well-known products, other than meat or produce, on which manufacturers' or distributors' allowances are given. Kroger stores passed along part of the discounts on these rogers to consumers in the form of lower prices; the remainder was retained, so that rollers THE KROGER CO. 723 639 Opinion despite their low price, were among the most profitable items sold on Kroger shelves.

To publicize its Everday Low Price policy, Kroger initiated the Price Patrol advertising campaign in the Atlanta KMA late in 1972. (RX 900 p. ) It spread to six other KMAs and, at its peak, covered 100 cities in fourteen states. (ID 104-105.) Each KMA ran the campaign according to instructions contained in manuals distributed by Kroger s general office. (See CX 15-16.) Kroger employees responsible for implementing these guidelines in their respective KMAs initiated the program by setting up teams of local housewives, who were instructed to make weekly comparisons of the prices charged on selected products by Kroger and its major competitors. (4) The products covered by the survey were selected by the Grocery Merchandisers in participating KMAs. They would pick approximately 600 items, which the KMA Advertising Manager would then group into four lists of 150 products or so each. (ID 71.) These lists were given on a rotating basis to Price Patrollers each week over the course of a three-month period, at which point the master list would be reviewed by Grocery Merchandisers for necessary changes. In their product selections, Grocery Merchandisers concentrated on items that were purchased frequently, were well-recognized by consumers, and were likely to be stocked by competitors. Also, Kroger offcials testified that selections were made from among all types of "groceries" in the store, in order to ensure that survey samples represented all product categories. (E.g., Tr. 1574-75.) This procedure was subject to some important limitations, however. First in Kroger s parlance, the term "groceries" does not cover fresh meat produce, or delicatessen (including fresh baked) items; these items were. in fact, systematically excluded from the surveys because Kroger perceived quality variations between such products sold in its stores and those sold by competitors. (Tr. 2101, 1574-75.)4 Second private label or house brand items, though within Kroger s defini- 2 ex IS, a Pri"", Patrol manual written in Kroger s headquarters and distributed to KMAs, contains examples of how the roiler system wurks. The fonowing table taken from page 4 of the manual, ilustrates the effed of a roller On the price and profitability of a 50 foot roll of aluminum foil. Figures in parentheses are evidently unit costs Regular Cost $12.82 (.256) "Roller " Cot $ 9.82 (.96) Reg-ular R"u,il $15.00 (:JOO) "Roller" Retail $13. (.270) Normal Pennies Profit $2. 18 (.044) "Roller" $3.68 (.74) Regular Mark-up % 14 RoHer" Mark-up % 27% Experience has shown sales on the above item almost doubled in a 4 week period giving an additional S500 mark- 2 On the average, changes to the lists were made twice a year. Reasons for modifications included the introduction of new prodm;ts into the market and changes in the popularity or availabilit.y or items. (1'r. 2127-28.) . At one point, Kroger apparently intended to offer evidence that some Price Patrols induded some meat items and product, Order Ruling On Complaint C(Jun. Motion for Summary Decision at 16 (May 17, 197H) However, the AI J ultimately found that fresh meat and produce, as well as delicatessen items, were not chosen for the Price Patrol survey lists. (ID 93.) In it. appeal, Kroger ha. not disputed this finding. (See RAE 4) g., Opinion 98 F.

tion of groceries, were generally excluded from survey samples again because of quality variations between Kroger and its competitors; milk and dairy products, which Kroger considered to be of equivalent quality in all stores, were included, however, along with a few other private label items.

To avoid biasing results, Kroger did not tell Price Patrollers which of its stores or its competitors' stores should be checked or on which day of the week the surveys should be run. (JX 2, Stec deposition p. 79.) Furthermore, all Price Patrol results were checked for accuracy by Kroger s advertising department. Items on which errors had been made to Kroger s advantage were eliminated from the survey; those on which errors favored a competitor were retained. (ID 86. After reviewing the Price Patrollers' checklists, the KMA advertising departments prepared "recap" sheets, which summarized the weekly survey results: Kroger s prices were compared with those of each competitor, and a table was (5)compiled showing the numbers of items on which Kroger was lower than, equal to, and higher than each other retailer in the survey. (RX 902; JX 2, Stec deposition at 90-91.) These tables or box scores were featured in weekly television or newspaper advertisements. Typically, the advertisements showed that Kroger had more lower-priced items than each of its competitors. However, on the rare occasion that Kroger "lost" a Price Patrol check, its advertising would so state. The grocery merch"ndising department would then conduct a survey of its own to determine whether price adjustments were needed to assure that Kroger prices were as low as the competition s. (ID 90. A number of the television advertisements mentioned that the weekly surveys compared "popular brand name items. (See, e. 537 pp. 1- ) Some commercials also specified that comparisons were limited to national or regional brands and that Price Patrollers never compare store brands against national or regional brands because this "wouldn t allow fair comparison." (RX 599.) However none of the television commercials explicitly disclosed that meat and produce had been systematically excluded from Price Patrol surveys. Similarly, the newspaper advertisements failed to disclose the exclusion of meat and produce until 1976, when announcements of survey limitations began appearing in relatively small print near the Price Patrol claims. (ID 116; RX 595.

The television commercials and some of the newspaper advertisements did invite consumers to inspect the Price Patrollers' checklists, which were regularly posted in Kroger markets. These checklists showed which items had been surveyed and what prices had been charged. The record indicates that a significant number THE KROGER CO. 725 639 Opinion consumers consulted them; one Kroger study found that between 8% and 30% of food shoppers who associated the Price Patrol campaign with Kroger had referred to the checklists in the stores. (CX 708 p. Kroger KMAs increased their advertising budgets to mount the Price Patrol campaign and placed their emphasis on television commercials, although Price Patrol results were also announced in large newspaper advertisements. The campaign was run in some KMAs for nearly six years, compared to a lifespan of three or four months for most of Kroger s other advertising programs. (Tr. 2093-1.) The evidence suggests that the campaign effectively informed consumers of Kroger s Everyday Low Price policy. (6)Consumer research showed high levels of public awareness and credibility in areas where the Price Patrol had been run. (CX 708.) Furthermore Kroger sales, profits, and market share increased significantly during the Price Patrol campaign. In Atlanta, for example, Kroger market share doubled. (Tr. 1870 2159.

The Price Patrol program was terminated in all cities by May 1978. Kroger states that its decision to end the program was prompted primarily by the fact that the present case had been brought. (Tr. 2130, 2160.

Kroger ran two other comparative price survey programs during the time the Price Patrol was in effect. Both were designed specifically to enable the company to assess its price competitiveness and their results were never advertised. The first survey, referred to as the Full Book Check, was conducted by Kroger employees at least once a quarter and in some areas as often as every six weeks. The check covered the prices of 2 000 to 4 000 items, sometimes including meat and produce. Results were tabulated in terms of the number of items on which Kroger was higher than, equal to, or lower than particular competing stores. (Tr. 2342--7.) The second of these surveys was known as the Burgoyne check, after the name of the firm that conducted it for Kroger. This was a market-basket survey-a survey of items selected to represent the types of purchases a typical consumer might be expected to make. The products surveyed regularly included meat and produce, as well as grocery items. Kroger employees selected the survey items and attempted to ensure that their choices would produce a sample that reflected the proportions of a consumer s food bil attributable to . One KMA, evidently typical of ot.her pent sixty percent of its advertisinf: funds On television time and twenty pen;"ot for newspaper space. (Tr- 1712) Only fifteeo perceot of the newspaper advertisioK space coocern"d Pric.. Patrol resultsthe rest was given to promoting speific Krocery, meat, or produce items- Kroger therefore ert that only 5% of the advertising funds devot.edu) Price Patrol went to newspaper ads, while 95% of this budget was used for television commercials. (RAE at 22, fn. 33, ) However, it should be noted that in the majority of newspaper advertising, the spal:e set aside for the Price Patroi was highly visible; for example, the Price Patrol surveys Were usually referred to in th..advertisements' headlines. 726 FEDERAL TRAm; COMMISSION DECISIONS Opinion 98 F.

meat, produce, and grocery items. Like the Full Book Check, the Burgoyne check was conducted on a quarterly basis. (Tr. 2321-30. DISCUSSION Summary Judgment on Meaning of Advertisements On May 17, 1978, the ALJ issued an order ruling on a motion by complaint counsel for summary judgment. The law judge denied part of the motion but agreed with movant's assertions that the Price Patrol advertisements had made the following claims: 1. The results of the Price Patrol Survey prove that most items in respondent' s stores are priced lower than in competitor s stores. 2. The results of the Price Patrol Survey prove that shopping at Kroger, rather than at competitors' stores, will result in lower overall expenditures. (7) 3. Shopping at Kroger, rather than at competitors' stores, will result in lower overall expenditures.

4. The Price Patrol Survey is a methodologically sound survey. Kroger opposed the summary judgment motion, and it reiterates on appeal two of the arguments it presented below. First, it argues that summary judgment may not be granted to resolve wholly factual issues such as the meaning of its advertisements, but is appropriate only for resolving legal questions where no factual issues are in dispute. Second, the company asserts that complaint counsel have failed to carry their burden of proof because they introduced no extrinsic evidence to support their reading of the Price Patrol campaign. These arguments were rejected by the ALJ on the authority of Ford Motor Company, 87 F. C. 756, 794-97 (1976). Thus Kroger now raises a third argument: that the ALJ misread Ford which respondent says was not intended to authorize a law judge to ignore federal judicial precedent on summary judgment procedures. We agree with respondent' s interpretation of Ford. We have previously applied our summary decision rule, Rule 3. , consistently with case law construing the equivalent provision of federal civil procedure, Rule 56. Lehigh Portland Cement Company, 78 F. 1556 , 1557 (1971); The Hearst Corporation 80 F. C. 1011, 1014 (1972). The Ford opinion signals no departure from this policy. The decision does establish that summary judgment motions may be granted in advertising cases. However. it provides little guidance as , Se ID 109- 112. which restate the AI..' s conclusion that the adverti"ements made these claims. THE KROGER CO. 727 639 Opinion to when summary decisions concerning advertisement claims should be issued and the burden of proof imposed on movants. These issues must be decided, as Kroger asserts, on the basis of accepted principles of summary judgment procedure. (8) As to its first argument concerning the propriety of a summary decision on the meanings of advertisements, Kroger is clearly right that a motion for summary decision is an inappropriate vehicle for resolving genuine factual controversies. Just as clearly, however such a motion may properly test the genuineness of asserted factual issues, and the ALJ may issue an order "specifying the facts that appear without substantial controversy." Rule 3.24(a)(5). Kroger real argument in this regard, however, seems to be that an order of the sort described in subsection (a)(5) may be sought only if the judge has denied a motion on all or some of the ultimate claims in a case. There is certainly precedent for the denial of summary decision motions whose only goal is to adjudicate questions of fact that do not dispose of any claim. See Yale Transport Corp. v. Yellow Truck Coach Mfg. Co. 3 F.RD. 440 (S. Y. 1944); 6 Moore s Federal Practice n56.20(2. 2j. However, in federal practice, facts that are not seriously disputed may be disposed of at a pretrial conference under Rule 16. Professor Moore indicates that it is preferable to resolve such matters under Rule 16, which leaves the decision to schedule a pretrial conference to the court's discretion, rather than under Rule , which requires a judge to specify the facts that are not disputed. Moore also notes, however, that distinctions between these rules have faded in districts where cases are pretried as a matter of course and observes that the Yale Transport court went on to dispense with the factual issues before it under Rule 16. 6 Moore s Federal Practice n56.20 (3-2) at 1216.

We believe that Kroger s objection to the procedure followed by the ALJ here lacks substance. Commission Rule 3.21 provides for the simplification and clarification of issues in pretrial conferences and makes at least one such conference mandatory in every case. Rules 21 and 3.24 both equip ALJs adequately to dispose of factual issues before a hearing if they are not genuinely disputed. Kroger was not injured in the least by the ALJ's decision to grant, in part, complaint counsel' s motion under Rule 3. , rather than to require them to resubmit under Rule 3.21. See Yale Transport Corp., 3 F.RD. at 441 1 In Ford the parlies moved for summary judgment on all issues in the case; hen' , the AI I'H summary decision concerns only factual matters, a distinction that Kroger contends is critical. l"urthermore, the question faced in Furd was whether a summary judgment motion On advertisements should be granted when the party opposing it had offered to submit Consumer surveys that it claimed disputed the existence of the claims alleged by movllnts. Here, Kroger h"" not offered any ext.rinsicevidence of iL Own On tbe proper interpretation of its advertisements. The company charges that it had no need to do so given the impropriety of the decision and compJaintcounsel' s failure to carry its own burden of proof. ), Opinion 98 F.

(after concluding that motion could not be properly brought under Rule 56(d), court decided to permit it under Rule 16).8 Accordingly, we (9Jhold that the ALJ's decision to expedite proceedings by disposing of uncontroverted facts prior to trial was not improper. Respondent' s second argument is that complaint counsel have failed to carry their burden of proof. The law is clear that, on a motion for summary judgment, the court must draw all factual inferences against the movant, and may rule in the movant's favor only if persuaded beyond doubt that no genuine issues of material fact exist. Adickes v. 8.H. Kress & Co., 398 U.S. 144 (1970); 6 Moore Federal Practice n56. 15(3j at 466-7. '" Kroger asserts that the claims alleged by complaint counsel are not within the literal meaning of the Price Patrol advertisements, and that complaint counsel have not even attempted to support their interpretations with extrinsic evidence. According to Kroger, the existence of the purported claims thus remains a matter of factual inference, and that Kroger is entitled to have that inference drawn in its favor because it has opposed the summary decision motion. It is settled that the Commission has sufficient expertise to determine an advertisement's meanings-express and impliedwithout necessarily resorting to evidence of consumer perceptions. National Dynamics Corp. 82 F. C. 488 , 548 (1972), aff'd 492 F. 1333 (2d Cir. cert. denied 419 U. S. 993 (1974). This is not to say that an advertisement is susceptible to every reading that it may technically support, no matter hf'w tenuous it might be; rather, the interpretation must be reasonable in light of the claims made in the advertisement, taken as a whole. National Dynamics Corp., supra; Firestone Tire and Rubber Co., 81 F. C. 398, 455 , 457 (1971), aff'd 481 F.2d 246 (2d Cir. 1973). In many cases, the Commission has refused to accept particular interpretations urged by complaint counsel because the advertisements themselves did not imply them and no extrinsic evidence had been offered to prove their apprehension by (10Jsome reasonably significant number of consumers. E.g., . Procedural!y, both a pretrial ruling under Rule 3.21 ami a partial summary adjurJic"tion under Rule 324(a)l5) are interlocutory in nature. Thus, regardless of which proct'urt' is followed, a litiganl would have the same right to sek interlocutory Commission review of the AW's determinatiun. See Rule 3.23(b) . We note that complaint counsel'smotion din in fact seek a summary determination that the advertisement's failure to disclose the exclusion of meat and produce from the Price Patrol surveys constituted a violation of Section fi of our Act. The law judge denied this portion of the motion because Kroger proffered evidenceOmethat meat item, produce, and house hnmds were included in certain surveys. Order Ruling on OJloplaint Corw.sel's Motion for Summary Decision at 16 (May 17, 1978). " The traditiunal test applied by the courts has been that the movant must establish beyond the "slighte doubt that no genuine factual issue exists. Some circuit now require only that. the movant show there is no " reli! and substantial" duubt concerning the absence of genuine issues.See Moore s Federal Practice 56. 15l3J n.6 at 467 69 & Supp. ln ofar as this Cl1se i concerned, we believe that the same results should be reached under either test. We find it unnecessary, therefore !. choose between the standards of "slighte t." doubt and "real and substantial" doubt.

TJ-- KltUli.lH cu.

639 Opinion Pfizer, Inc. 81 F. C. 23, 59 (1972); Sears, Roebuck and Co. 95 F.'l. 406 510-11 (1980). n Nevertheless, we do not agree with Kroger s apparent argument that an implied claim can never be determined on a summary decision motion absent the support of extrinsic evidence. Applying accepted principles of summary judgment law to the case at hand seems to us that the critical issue is not whether the alleged claims are implicit, but simply whether they are so clearly conveyed by an advertisement that no genuine issue as to their existence can be raised. Where such certainty exists, the movant may be said to have fully discharged its burden of proof under Rule 3.24." (ll) We now turn to the particular claims found by the law judge to determine whether they were stated or implied by the advertisements clearly enough to justify a summary adjudication. The claim that most items in Kroger stores are priced lower. The respondent contends that its advertisements stated merely that Kroger had more items with lower prices than its competitors not that most items in its stores were cheaper than those sold by competitors. OUT review of the advertisements in the record bears out Kroger s assertions. The newspaper advertisements cited in " When extrinsic evidence On the meaning of an adv('rtisement has ben introdueed, it Inust, of course, he considered hy the law judge ..nd the Commission, SeeCinderella Career Finishing SchrJ(L. I",;. v. FTC 425 Y2d 583, 588 (D.C. Cir. 1970). Such evidence does not wholly supplant Commission expertise, however, but simply contributes to the Commission s ability to form a comprehensive understanding of the ..advertisement.Firestone Tire and Rubber Cu. supra 81 F.TC. at 454; Crown Central Pelmleum Corp. 84 F, C. 1493, 1540 (1974). An advertisement may convey 3 variety of meanings, depending not only On the complexity nf iL text and method of presentation, but also on the variation in backg"rounds and interests uf its audience.ContinentalSee Wax Corp. 330 F.2d 475 (2d Cir. 1964);Murroy Space Shoe Corp. v. 1'7'(.' 304 F.2d 270 (2d Cir. 1962) Thus, consumer survey evidence suggesting the predominance of speific claims does not nece arily indicate the nonexistence of other implied representations. Of course, the challenged claim must be a reasonable interpretation of the ..advertisement, but it is scttled that an advertisement capable of several reason..ble interpretations m..y violate Section 5 if but One oftbom is deceptive or unfair.Chrys/erCorp. v. 561 F.2d 357, :i6:J (D.C. Cir. 1977);Res"rl Car Renlal Syslem, Inc. v. FTC 518 F.2d 962, 964 (9th Cir- 1975);C"nlinental Wax v. YI'C, supra; Murroy Space ShoeCrJrp. F'I'C sopro " Once a movant has made a satisfactory prima facie showing of the absence of genuine factual issues, the opposing party bears the onus of resurrect in I' the possibility ofa dispute concerning material fact, which it may do by fiing affdavits that set forth appropriate fact Or that explain its present inability to state the facts justifying opprmition to the motion. Rule 324(a)(4). Se alsoMnurning v, Family PublicaliunB Sproice, Inc. 411 U.s. 356 (1973); Adickes v. s,II Krpss Co. 398 U.S. 144 (1970); 6MarJre " Federal Prodice 1156.15l3J at 48:-17. The opponent is not entitled to hold back evidence it would have relied on at trial, nor may it forestal! summary judgment by asserting immaterial facts or setting forth merely speculative argument . See 6 Moore s Federal Pradice 56, 15l3J at 485-7, and cases cited therein. Here Kroger states that it "offered no evidence on the meaning of the commercials" in opposition to complaint counsel's motion, but argued primarily that complaint counsel had failed to demonstrate the absence of a triable issue. (RRB at 24, fn, 26.) As to those claims fur which cumplaint counsel have met their burden uf proof, Kroger cannot now complain that it was deniedthe right tn prescnt extrinsic evidence nn the pi-oper interpretation of its advertisement.5 or that the law judge may not apply his expertise "in a vacuum." (See RRB at 26.) If Kroger had facts suggesting that complaint counsel's reading of the Price Patrol commercials was wrong, those facts should have been presented in opposition to the summary judgment motion. 361- 444 a - 82 - 47 QL 3 Opinion 98 F.T.

complaint counsel' s motion carried such headlines as Kroger Shoppers Save More-The Price Patrol Proves It.'''" and " Price Patrol Proof-Kroger Leads with Low Prices in Atlanta!"" The advertisements' copy often stated:

Kroger is dedicated to the principle of pricing products so that we will be lower on more items more often than anyone else." We have lived up to that pledge as proven by the Price Patrol 15 Kroger is lower on more items than any competitor the Price Patrol checked. . Statements in the television commercials are to the same effect. These are undeniably forceful claims, ones that clearly promise that shoppers will save money by patronizing Kroger stores. But we do not think that the advertisements either state or imply that most items in Kroger stores are cheaper than in competitors' supermarkets. Nothing in the newspaper advertisement headlines or in the text of print or television advertisements makes such suggestions. Moreover, while the (12Jbox scores used to publicize Price Patrol results usually showed that Kroger had more lower priced items than each of its competitors, they also usually showed that the number of such items was less than the total of items on which Kroger s prices were equal to or higher than those of its competitors. Thus, we conclude that there is a genuine factual issue about whether a "most items" claim was communicated by these advertisements. It was therefore improper to issue a summary decision holding that the claim was made, and we reverse the ALJ' determination on this point.

The claim that Price Patrol surveys prove that shoppers save overall at Kroger and that surveys are methodologically sound.

The advertisements in the record are quite sufficient to demonstrate that Kroger claimed its Price Patrol results proved that shoppers would save by patronizing Kroger stores. As the quotes in the preceding paragraphs demonstrate, Kroger s newspaper adver- " Complaint couns,,'s f' irst Request for Admissions, Attachment :J(k) " Id. Attachment 3(d); see also RX 551-556. (RXs 552 and 555 appear in Appendix. " Id. Attachment-; 3(d) and (g); Se" also RX 540. (RX 540 appears in Appendix. '0 Id. Attachments 1(i), (r), (u) " It would, of course be appropriate to remand this issue to the AW for receipt of further evidence. However as we discu below, other evidence in the record independently justifies forbidding respondent from unfairly or inaccurately advertising- the results of comparative food price surveys. Consequently, it is unnecessary to determine whether Kroger s advertisements actually made the " most items" claim and, if so, whether that claim was substantiated. SeeSears Roebilck ond Co. 95 F. C. 406, 512 (J9ROJ, petition fiw review (il No. 80--7368, 9th Cir. July17 1980 " A fortiori we find that Krog-er also made t.he basic claim that consumers would save overall by shopping at its stores.

. . , p. UJ! .tc\VU'-H 639 Opinion tisements could hardly have been clearer in representing that the Price Patrol uproved" shoppers save more at Kroger. The newspaper advertisements are sufficient in themselves to establish that the " proof of overall savings" claim was made. However, we find that Kroger s television commercials, though more muted in tone, also represented the Price Patrol results as proof that one would save money by shopping at Kroger. A script from one of the commercials, introduced by respondent and found to be typical by the ALJ, should illustrate the point. It reads in part as follows: 1. The best way to find out who really has the low prices in the Atlanta area is to check the leading food retailers each week. 2. That's what the Kroger Price Patrol does. 3. We checked prices on 150 popular name brand items, and found Kroger lower on this many items. higher on this many items. and the same on this many items.

4. Once again, Kroger was lower on more grocery items than any other store we checked. (13) 5. Read our complete report at Kroger. See why we believe Kroger is doing what it takes to stay the Atlanta low price leader. (RX 537 Unlike the newspaper advertisements, this script does not contain the words "proof' and Usavings." However, they are conveyed emphatically enough by the advertisement as a whole. The attention given in the script to Kroger s low prices conveys to consumers a message that they can save by shopping at Kroger; the viewing audience is doubtless interested in Kroger s pricing policy chiefly because the information may help them stretch their dollar further. The overall savings claim, then, is quite clear. It is also clear that the television commercials used the Price Patrol surveys to demonstrate reliably that shoppers would save overall at Kroger stores. The advertisements attempted to back up the overall savings claim by featuring the Price Patrol survey data and by advising viewers that these surveys were "the best way to find out who really has the low prices.. . ." The Price Patrol surveys were obviously presented not merely to show that Kroger was lower than its competitors on the items checked, but also to demonstrate that Kroger would be cheaper overall. The commercials invite consumers to assume that the Price Patrol data can be projected reliably to all types of items sold by Kroger. Consequently we conclude that the television commercials imply strongly what is explicitly stated in the newspaper advertisements-that the Price Patrol proves shoppers would save overall by patronizing Kroger. We also conclude that both the television and the newspaper advertisements represented the Price Patrols as being methodologi- Opinion 98 F.

cally sound. The representation was implicit in Kroger s claim that the surveys were proof of Kroger s lower prices. Inherent in the proof claim is the notion that the surveys were conducted in such a way that their results are reliable enough to prove the overall savings representation. This is simply definition of methodological soundness. Thus, because Kroger did claim its surveys had provided proof, it also implied that they were methodologically sound. OUT approach in interpreting Kroger s television commercials has not been to focus only on what words are spoken; rather, it has been to follow our established policy of identifying the advertisement' general impressions. Standard Oil Company of California 84 F. 1401 , 1472 (1974), modified on other grounds, 577 F.2d 653 (9th Cir. 1978); Rhodes Pharmacal Co. v. T.C. 208 F.2d 382, 387 (7th Cir. 1953). Kroger argues, (14)however, that any claims of " proof' or methodological soundness" are belied by the overall "folksiness " of its television advertisements, and seeks to draw a paranel with Pfizer, Inc. 81 F. C. 23 56 (1971), in which the frivolous nature and aura of sexiness" in a challenged ad apparently convinced the Commission that Pfizer had not impliedly represented that it had scientific tests to substantiate its performance claims. The folksiness cited by Kroger is undeniably present in its commercials-Price Patrollers introduce themselves by their first names and informally begin some of their commercials by explaining that a colleague s absence is due to the fact that she just had a child or that her husband was transferred to another city. Still, we do not find that these characteristics detract from the significance or perceived reliability of the survey results. Furthermore, Kroger commercials are quite distinct from the advertisement litigated in Pfizer. The Pfizer advertisement concerned a sunburn ointment which, according to the blond (female) television model, actually anesthetizes nerves and stops sunburn pain in Hless time than it takes me to slip out of my bikini." 87 F. C. at 59. Nowhere in the commercial did the advertiser suggest it had conducted any tests to support these claims. See Id. at 58 nA. Here, the situation is wholly different since Kroger built its entire advertising campaign around the surveys conducted by its Price Patrols. .. Kroger has alsQarguI'd that, given th.. l10nscientific nature of its commmdals, the Price Patrol 511rv"yS could have been viewed as providing orily the kind of proof on which "housewiv..s would rdy or on which business (WapI.. would rely in making decisions. " (RAB 1'.24.) There is certainly nothing lo this effect suggested by the advertisements themselves. Moreover, Kroger has failed throughout these proceedings to suggf's( what specialized kinds ofproc)f it bdieve that housewives and executives Cllstomarily f"ly on. We are satisfied commonthat sense interpretation of the Prin' Patrol campaign supports complaint counsel's position that the surveys were repmsentcd as providing "prool" a.,; that term is gefJer l!y understoo that shopping at Kroger will res\I!t in savings. It was incumbent on Kroger to proffer evidence suggesting that this readinf: uf the advertisements w iflcurrecL Se note 12,upr-. Kruger chose not to do so, howf'ver THE KROGER CO. 733 639 Opinion Both the television and newspaper advertisements leave us with little doubt that Kroger has presented its Price Patrol surveys as methodologically sound and as proof that consumers will save at Kroger stores. Accordingly, we affrm the ALJ's decision regarding these claims. (15) Adequacy of Substantiation It is now well-settled that Section 5 of the Federal Trade Commission Act prohibits advertisers from making affirmative product claims that they have no reasonable basis to believe are true. Pfizer, Inc., supra; Firestone Tire and Rubber Co., supra; National Dynamics Corp., supra; National Commission on Egg Nutrition, 88 C. 89 , 174 (1976), modified on other grounds 570 F.2d 157 (7th Cir. 1977). Such practices are unfair because they materially impede a consumer s ability to make an economically rational choice, and a competitor s ability to compete on the basis of price, quality, service or convenience. Pfizer, Inc., supra 81 F. C. at 62. See also Commission Statement of Policy on Consumer Unfairness in Commission letter to Senators Danforth and Ford, Dec. 17, 1980. They are also deceptive because (mpany consumers are likely to assume that when a product claim is advanced which is in theory subject to objective verification, the party making it possesses a reasonable basis for so doing and that the assertion does not constitute mere s part." Nationalsurmise or wishful thinking on the advertiser Commission on Egg Nutrition, supra 88 F. C. at 191; cf Jay Norris Corp. , et al. 91 F. C. 751 , 855 (1978), aff'd 598 F.2d 1244 (2d Cir. 1979).

There is no bright line test for deciding what quantum of evidence is needed to provide a reasonable basis. These determinations are best made in light of the particular facts in each case. However, the Commission has noted that the following factors are relevant: (1) the type and specificity of the claim; (2) the product involved; (3) the possible consequences of a false claim; (4) the degree of consumer reliance on the claim; and (5) the type and accessibility of evidence to support the claim. Pfizer, Inc., supra, 81 F.'l. C. at 64. Citing these factors, the ALJ concluded that Kroger s Price Patrol surveys were not designed and conducted according to procedures recognized as sound by survey experts, and, notwithstanding the company s advertising, failed to prove that shoppers would save overall by patronizing Kroger stores. He consequently held that Kroger had no reasonable basis for those claims and had advertised them in violation of Section 5. (ID at 76.) Kroger disagrees with this analysis of its surveys. It contends that, given the special difficulties Opinion 98 F.

posed by retail food price surveys, its Price Patrols were reasonably sound and furnished adequate support for the savings claims that consider thesethe ALJ determined had been made. We will contentions presently. First, however, we must assess Kroger threshold argument that the reasonable basis doctrine should not even be applied to comparative food price claims. (16) Kroger asserts that the reasonable basis doctrine arose in the contexts of health and safety claims and complex, technical performance claims that could be verified by the manufacturer through a single test. It contends that such claims are distinctive, and that the rationales for requiring that they be reasonably substantiated do not apply in the case of comparative food price advertising. Health and safety claims, for example, warrant substantiation in view of the risks to which false representations would expose consumers. Here no such risks are presented. Technical performance claims cannot be readily verified by consumers, whereas the advertiser can typically test them without undue expense. Here, Kroger s weekly survey claims would require frequent substantiation--ven though, according to Kroger, shoppers could easily determine for themselves which supermarket is cheapest and market forces would discipline those stores whose price claims were false. (RAB 41. We do not find these arguments convincing. However one may characterize the claims in past substantiation cases, the Commission s opinions clearly indicate that the reasonable basis test was not confined to claims affecting health or safety. The Pfizer decision referred simply to "affirmative claims " 81 F. C. at 62, and in National Commission on Egg Nutrition the Commission suggested that substantiation should be obtained for product claims capable in C. at 191; see also Jay Norristheory of objective verification, 88 F. Pfizer thatCorp., supra 91 F. C. at 854. The Commission did note in some claims may not require substantiation, but it spoke in that regard merely of puffing and hyperbole. It contrasted those subjective forms of advertising to specific product or service claimsprecisely the kind of claims disseminated by Kroger in its Price Patrol campaigns. 81 F. C. at 64-65.

Moreover, we do not accept respondent's view that consumers can easily tell which store is cheapest. To make that determination accurately-and without relying on unsubstantiated advertising claims--onsumers would have to shop comparatively each week among several stores over a significant period of time, an effort involving far greater inconvenience than it would be reasonable to 20 It is true that the reasonable basis doctrinedemand of consumers. W," do note, however, that the harm to consumers in this instance is not a. g"real as it might be in cases (Continued) (;j THE KROGER CO.

639 Opinion would require that Kroger s weekly food price survey claims be subject to repeated substantiation, but this fact alone does not suggest that Kroger s advertisements should be exempted from the reasonable basis doctrine. Rather, it raises the more complicated question, to which we now return, of whether Kroger s survey data did, in fact, adequately substantiate its Price Patrol claims. (17) The ALJ determined that the Price Patrol surveys were methodologically unsound in several respects. First, he found that although the surveys systematically excluded meat, produce, house brands and certain other items, the advertisements did not clearly disclose these omissions (ID 114, 178.) Second, the ALJ found that, even among those product categories not excluded from the survey lists Kroger s sampling procedures were deficient. He observed that samples were not drawn randomly, so as to supply a basis in probability theory for projecting sample results to a larger population of items. Survey samples were instead drawn by the same Kroger personnel who were responsible for implementing the company s Everyday Low Price policy. Relying on the testimony of both complaint counsel's and respondent' s experts, the ALJ concluded that this combination of functions rendered the surveys unsound because it unreasonably increased the possibility of bias. (ID 183- 187.) As to the Price Patrol samples themselves, he found that they were not representative because they were not weighted to reflect the frequency with which an average shopper bought certain foods and because they contained disproportionately large numbers of rollers and high-volume, fast-moving items. (ID 179-J82, 191-193. Lastly, the law judge determined that the advertised charts would not support any conclusion about whether consumers would save money shopping at Kroger. The flaw in the charts, according to the ALJ, was that they failed to show the differences between Kroger and competitors' prices, and compared only the number of items on which Kroger had prices lower than its competitors--r vice versa. The ALJ found, however, that it was possible for Kroger to have had more lower-priced items than any other store and still be the most expensive overall. lid 188-190.

Kroger contends that the standards to which its surveys were held by the ALJ are unjustifiably strict. It asserts that no one has disputed the basic claim that its prices generally were as low as or lower than those of the competition. And while it admits that the where daims are impos. ibe to verify prior to purchase, and the p"tentia! economic tOEEis high or there arc significant health or safety risks associated with the use of a prlluct. As previously not, the nature of the injury resulting from an I.Hlsubstantiated claim is an appropriate consideration in determining what kind of substantiation is ca!jed for. PfizerSee 81 F. C. at 64. Opinion 98 F.

Price Patrol surveys were not perfect, it also claims that they reflected a sensible compromise between the ideals of statistical theory and the restraints imposed by limited financial resources. Kroger says that the Price Patrol surveys were comparable in quality to the kinds of surveys that businesses frequently rely on for long-range planning. Moreover, it argues that if the AM' s tougher survey standards are upheld, comparative retail food price advertising wil disappear, since no company wil be able to afford the necessary substantiation. The short-term result, according to Kroger, is that consumers will no longer receive useful information on competing stores' pricing policies- That, in turn, will injure discount stores, who wil be unable to advertise their lower prices; and by the same token, it wil benefit higher-priced merchants, who can continue to attract business by touting superior service, selection, or other non-price attributes. (18) The Commission recognizes the importance of these concerns. have endorsed truthful comparative advertising" and do not intend to stifle it indirectly by requiring unneeded substantiation. The record in this case does not reflect how much the cost of respondents Price Patrol campaign would have risen had the company s surveys fully matched" those described by complaint counsel's expert witness. Nevertheless, we are willing to assume that the increase would have been substantial Furthermore, we agree that the evidence does not dispute Kroger s basic argument that its prices were frequently below or equal to those of its competitors. See 925-964. If respondent' s advertising had actually been confined to making correspondingly limited claims, the data that it had the Price Patrol and the Full Book and Burgoyne checks-would have come much closer to providing adequate substantiation. As we have found, however, the Price Patrol claims were considerably more specific, and consequently, a good deal stronger as well. Respondent' s advertising represented not only that Kroger was the low price leader, but also that it had run surveys which proved on a weekly basis, consumers would save by shopping at its stores. The point was driven home by the tables in the advertisements showing the numbers of items on which Kroger s prices were below those of its competitors. In effect, Kroger made its own substantiating data an integral part of its advertising. Consequently, consumers and competitors may fairly expect that the substantiation wil, in fact, closely match the representations made about it. See Litton 21 Commission Sl.l.f''nentofl'olicy in Regard to Comparative Advertising (June 27 1979). 22 Although we uitimately find respondent' s substantiation to be inadequate, we do not go as far as complaint CQUllscl or their expert witness ill terms of the sljrvey procedures that Kroger must follow in the future, , THE KROGER CO. 737 639 Opinion Inc. No. 9123, slip op. at 7 (Jan. 5, 1981), petition for review docketed No. 81-7148, 9th Cir., Mar. 12, 1981. It is thus inadequate for Kroger to say only that its surveys were a reasonable compromise between statistical ideals and financial realities, or that business executives would have relied on them for their day-to-day decisionmaking. Of course, even for proof-type claims, as we have here, some compromises in survey techniques may be appropriate, but the issue should not turn exclusively on cost or on whether the results would be acceptable to business executives, who undoubtedly are in a better position than consumers to understand the limitations and deficiencies of a survey. Rather, the issue should center on consumer expectations and the reasonableness of the procedures in light of those expectations and the other factors discussed in Pfizer, supra 64.

We emphasize that the best possible survey technique may not be required for this purpose. We recognize that in most instances it will be difficult to observe every relevant statistical nicety. The elements of a sound survey methodology may be expected, therefore, to vary according to the type of the (19Jclaim and the product or service involved. Stil, this hardly means that every departure from statistical theory is permissible so long as it is dictated by cost. There are, we expect, certain statistical principles that may not be compromised without undermining a survey s reliability.23 Where the demands of the purse require such compromises, the advertiser must generally limit the claims it makes for its data or make appropriate disclosures to insure proper consumer understanding of the survey s results.

In identifying those elements of survey design and implementation that are critical to the reliability of a food price survey, we have paid close heed to the three experts who testified in this case. The expert testimony adduced by complaint counsel and respondent has convinced us that the Price Patrol surveys were not methodologically sound and thus did not prove that shoppers would save overall Kroger. Respondent' s experts, Dr. Benham and Mr. Oliver, did state that the Price Patrol surveys offered "useful" and "adequate comparative food price information when considered in light of other available data, such as the Burgoyne and Full Book survey results. .' Dr. Kohouttcstifjed We have certain ideals that are derived fromftheJ philosophy of science and other sciences, and any study that s done in the rea! world would usually involve compromises with certin part of that idea!.

There are, however, established standards for what compromises are acceptable and which Ones are not and if a study uses unacceptable compromises, then 1 wnuld say we could call that study unsound- (Tr. at 1004- 05.) " Mr. Olivertcstified On eross-examinationas follows Q. Well, let me see if I can restate the question- Would you recommend that in part as a consideration (Continued) Opinion 98 F.

These endorsements support Kroger s view that it had a reasonable basis to believe its prices were generally as low as, or even lower than, those of most competitors. But they fall short of confirming what Kroger s advertisements actually claimed-that each Price Patrol survey won by Kroger proved it had the lowest prices around. (21)Furthermore, respondent's experts based their opinions of the Price Patrol partly on the fact that the Burgoyne and Full Book data generally corroborated Price Patrol results. However, it seems clear to us that these surveys cannot be used to substantiate claims that the Price Patrol was methodologically sound or that its results were tantamount to proof. Kroger did not rely on the Burgoyne and Full Book data to substantiate these claims. (ID 206, 223.) Nor could it have. The Burgoyne surveys and Full Book checks were not conducted in many of the cities in which Price Patrol advertisements were run (Tr. 2429, 2785) and, in any event, were conducted too infrequently to provide adequate support for the weekly survey claims. Most importantly, Kroger "lost" to other stores In Burgoyne and Full Book surveys significantly more often than it did in the comparable Price Patrols. For example, in its Atlanta advertising campaign, Kroger claimed it had won every Price Patrol items be selected for a retail foo price survey beause they would likely be available at cornpctit"r ' stofes A. Yes, I think that should derinitely be a consideration. Q. Now, did I understand your testimony to be that the Price Patrol methodology, done th., way you understand it to be done, was adequate demonstration of the fact that Kroger was lower overall in all categories of foo?A It(20)was an adequate demom;tration of the geneml pricing structure of the chain comp,Hed to other chains on items that you cuu!d compare Q. Was it an adequate demonstration of the fact that Kroger was luwer uverall in ;III cMcgories of food? A. Mr. Correia, you have a problem here. These arejudgmpnt samples. You try to take something you can compare and draw conclusions from. If that information is consistent with the information on the other products, then thatseemstobea reasonable sort of judgment sam pI etodo Mr. Correia I move to strike that answer as nonresponsive Judgellyun' Motiondenied Q Are you saying that you need to look at other information besides the Price Patrol tu make that ddermination A 1 think you .'hnuld have other inf"rmation at your disposal, Mr. Correia (Tr.278:r-u.

Dr. Benham tpstified to similar dfcd on direct examinati"n The Wit.ness Again, as I stated in my previous response, the infurmati"n that the f,rm had internally to do its own monitoring, to make its own judgment about these matt.ers, impre""e" me as being very useful information on which re.:sonable individuals will make a reasonable determinatiun that, in fact. t.he stated pulicy of Kroger and the stated con.'equences of shopping with Kroger had a strong likelihoud uf being the adual.,experience of Consumers inshopping at those stores. So--. I will stop there (Tr. 2875; s,'e al.') Tr. 2865.

" Tr. 2786-2787 (testimony of Mr. Oliver). Dr. nenham did not testify specifically on whether Price Patrol information would be ad.,quate if it. were unsupported by Burgoyne and t' ul! Hook figures. !lis testimony, however indicates that he always took account of the latler set of data when as sing Krogcr luw price claims. See Tr. :'H74 28H3-2884.

Of, The Burgoyne surveys were typically run on a quarterly basis and Full Book checks conducted unce very four I." five weeks un the averag". (lD 224, 208.) A fonner Kmger grocery merchandiser testified, however, that each week the prices on 500 to 700 grocery it..ms in rpspondent's stores w"re chan ed (Tr. 165\)) lie also testified that the Burgoyne results, at least, were outd..ted by t.he time they were rec..ived. (Tr. 1658.) Dr. K"hout stated his view that the monthly Full Book checks were likewise insuffcient to support part.iClJ!ar Price Patrol re 111L (Tr 1D5Y--0.

Tt1.1 n.nv....

639 Opinion survey between 1972 and early 1978. Yet, of twenty-nine Burgoyne checks run during the same period, Kroger finished second or worse ten times; of nineteen Full Book surveys also run during that period Kroger was beaten by one or more competitors twelve times. (22) The qualified endorsement given the Price Patrol by respondent' experts, coupled with the unsuitability of the Burgoyne and Full Book data, cast doubts on whether the Price Patrol surveys are sufficient even as a supplementary form of substantiation. These doubts provide the backdrop to our consideration of expert testimony and other evidence on specific aspects of the Price Patrols that the ALJ found did not conform to sound statistical procedure. Exclusion of Meat, Produce, and Other Items Although respondent's advertising claimed that Price Patrol surveys proved Kroger was a cheaper place to shop than its rivals the surveys did not compare prices on fresh meats, produce, certain dairy products, and store-brand items. These exclusions represented a significant portion of Kroger s business: meat and produce alone accounted for 28% of Kroger s retail food sales. (ID 29.) Furthermore the Burgoyne and Full Book check results suggest that Kroger meat and produce prices were significantly higher than its other grocery prices, relative to the competition. For example, the Burgoyne checks conducted in the Dallas, Delta, Central, and Southland KMAs showed that Kroger s prices on grocery products other than meat and produce were undercut by another competitor 19 percent of the time. However, Kroger s meat and produce prices were bettered by a competitor in 86 percent and 82 percent of the surveys respectively. (ID 145; CX 251-282.

Kroger asserts that fresh meats, produce, and certain other items sold in its stores are of a higher quality than what its competitors offer, and that it should be permitted to drop them from a comparative price survey for that reason. (RAB 35.) We do not disagree. However, all three experts who testifled agreed that Kroger should at least have disclosed what types of products it had excluded from its surveys. (Tr. 1222, 2676, 2988-89.) This point strikes us as one of commonsense. If an advertiser systematically excludes a product from a survey sample, it should take pains to ensure that its advertising claims do not extend to that product, and " RX 925, 947 Similarly, Krof(erral1 an adve.rtisement in a June 1976 edition of a Chattanooga newspaper tating that it had lost Price Palrol survey for the. first time in 174 weeks. (RX 566.) Howev"r, during substantially the same p€riod, it was a runner-up in five- out of sl"tecn Burgoyne surveys. (RX 926.) Se also ex 8J:J.

g., , p.

Opinion 98 F.

it should therefore limit the scope of the claims by providing clear disclosure of the exclusion or by otherwise qualiying the claims. (23) Kroger maintains that it did make disclosures, and that it invited consumers to check the actual survey lists. To be sure, the television commercials, and to some extent the newspaper advertisements, did suggest to consumers that they:

check our report at Kroger. See why we believe Kroger is doing what it takes to stay the low price leader.

(RX 537 However, this invitation cannot itself discharge Kroger s responsibility to inform consumers of the survey s limitations. For one thing, the survey lists available at Kroger stores did not state that meat and produce had been systematically excluded. But even if they had disclosed this information, we do not think that Kroger s general invitation to "check our report" would have been adequate. Shoppers should not have been required to visit Kroger stores simply to learn facts that materially limited the Price Patrol claims, at least where Kroger could have limited its claims to begin with or could have made appropriate disclosures in the advertisements themselves. A number of the television commercials also stated that the surveys consisted of comparisons of "popular brand name items. (E. RX 537, pp. 1--.) And, the record contains some commercials that were devoted entirely to informing consumers that Kroger surveys checked prices on otnational and regional brands" and that: We check identical brands in each store. We never compare store brands against national or regional brands. This wouldn t allow fair comparison. (RX 535, p. 1.) These are helpful disclosures, but they are inadequate to dispel the notion engendered by Kroger s other television commercials that the Price Patrol results were applicable to all types of Kroger merchandise. While the televised disclosures mentioned that store brands were not matched against national or regional brands, they fail to reveal that store brands were excluded from the surveys entirely. The commercials arc similarly opaque about the exclusion of meat and produce. To be sure, the announcer notes that the surveys compared popular name brand products-which would exclude most meat and produce items-but given the sweeping claims made for the survey results, a more specific disclosure is necessary to ensure that shoppers understand the survey s limitations. This is particularly true here, where the survey results are represented as proof of an overall savings claim and where Kroger knew that its meat and THE KROGER CO. 741 639 Opinion produce were priced somewhat higher, relative to the competition than were its other grocery products.

Disclosures did begin to appear in at least some newspaper advertisements in 1976. For example, a full page advertisement contained the following notice: (24) Price Patrol checks only items which are totally or practically identical in each of the stores checked Fresh meat, fruits and vegetables cannot be accurately compared due to variations in grades, sizes, trim, etc. (RX 595.

The disclosure neglects to mention, however, that store brand items were not surveyed. Furthermore, it began to appear only toward the end of the Price Patrol campaign, and after the complaint in this matter issued. Even then, the disclosure was inadequate because it was inconspicuous and only intermittently used, Consequently, we hold that Kroger failed to disclose adequately that its surveys excluded fresh meats, produce, and certain other products, and that given the claims made for the survey results in its advertisements, this failure was inconsistent with sound survey procedures.

2. Sampling Procedures Complaint counsel's expert, Dr. Kohout, stated that survey samples should be drawn randomly, if at all possible, to minimize the chance that the conscious. or unconscious biases of those conducting the survey affect the results. (Tr. 1047.) He also favored the use of random sampling because it enables one to quantify how accurately the sample results may be projected to the population as a whole. (Tr. 1303.) It is thus possible to determine whether sample data provide statistically significant results; results that one may confidently conclude are not chance occurrences but reflect a definite pattern in the population. By contrast, Dr. Kohout said that judgment samples-samples selected by persons using their exper tise as to which assortment of items fairly represent the general population-leave room for the entry of personal bias and do not provide any basis for determining whether the data obtained are statistically reliable. (Tr. 1303-1304.

Kroger used judgment sampling procedures in developing its Price Patrol survey lists. While it agrees that random sampling is the preferable approach, it contends that its judgmental techniques were dictated by the peculiar problems of conducting comparative price surveys among retail food stores. The chief difficulty is that ,. Sec. , complaint counsel's First Request for Admissions, Attachments 3(p), 3(,,), :J(r), 3(5), all of which are newspaper advertisments disminate in latk 1976 and in 1977 without the disclosure. Opinion 98 F.

competing stores typically carry only some of the same brands of products. (See RX 1003, pp. 81 , 87.) Thus; if one draws a random sample ofthe goods carried in a particular store, there areexcellel1t odds that one wil find only a few of the same items on competitors shelves. Kroger asserts that this high degree of unavailability makes it extremely diffcult-perhaps impossible-to use random sampling procedures Jor food price comparisons, unless the (25)comparisons are made between non-identical products. Kroger believes that such comparisons could mislead consumers; who might not understand that price differences may reflect, to some degree, variations in the quality of grocery items carried by competing stores. The ALJ evidently accepted Kroger' s position on the impracticabilty of random sampling in the circumstances presented and did not rule that it was essential to a methodologically sound food price survey."O However, he did conclude that if judgment samples are used, those persons responsible for selecting the samples must not have a stake in the survey s outcome. That is precisely what occurred here. The Grocery Merchandisers who prepared the Price Patrol survey lists were also responsible for successfully implementing the Everyday Low Price policy. The ALJ held that the combination of these two functions was inconsistent with sound survey methodology. (26) Kroger contends that the only personnel who had the expertise to select judgment samples of food items were its Grocery Merchandisers. It states that it would have been prohibitively expensive to use an outside firm for the selection of survey samples because the firm employees would have required extensive, time consuming training to learn how to make appropriate sample selections. (RAB 38.) Since Kroger has cited no evidence to support this assertion, we cannot affirm its accuracy.31 In any event, all three expert witnesses agreed "'fheobje t Jesson for Kroger is the Commission s own Retail Food Price Survey (RFPS). This proj..ct wa.s first announced in 1973 as an effort to devise a survey protocol for comparing the overall cost of shopping at various supermarkels. 38 FR 1543 (Jln. 15, 197:-1 Staff began by compiling Ii randomly selected sample 'if grocery items that was suffciently large to permit statistically reliable projections of the reslIlts to the stores in general. To its dismay, thcstaff discovered that very few of the. sample items were available at the stores included in its test surveys, and subsequent attempts to control for the high incidence Df product unavailability proved unsuccessful. Ewmtually, the staff Dptedfora different sampling approach. Put simply, it used data from the U.8. Depurtment of Agriculture (USDA) to cla.sify aU grocery products into roughly 1 466 categories;it grouped th' se categories into stratabas on unit price, and it then instructe surveyNs toseled a product from each stratum in each store. According to the staff memoranda in the record, this approach was belleved to afford the most efficient means of obtaining statistically reliable price comparison among stores. Ilowever ooause the products selected in the various stores would not be identical, the approach would provid", consumers with no information ciJnccrning the quality of the goosin each store.(See RX lCKXI. ID 275-285.) '0 Complaint counsel did not appeal, so we presume they have '1diffculty with this determination. Since the suitability of judgment sampling is not an issl1e before lis, we expr",ss no opinion on when it may properly be used in lieu of random sampling.

"' The AL.J found that ihe aonuaJ cost of cooducting weekly Burgoyne surveys in 100 cities would be about $644 020. (ID 230.) Ashe noted, however, that estimate did not include the overhead cost! of the supervising Kroger employe€s' time and the costs of their office facilities. It also did not include the probable cost of training a staff of ICrmtifwedi THE KttUIJ.cn ,-,,.

639 Opinion that the function of sample selection should not have been combined with that of pricing. (Tr, 1037- , 2713, 2789, 2989-90.)Jt is not hard to understand why. One Grocery Merchandiser testified that after he prepared a Price Patrol survey sample he made a point of accepting roller" discounts on the products in the sample, so as to ensure that Kroger performed well on the surveys. (JX 1, Pelln deposition at pp. 58--1.) Such conscious efforts to influence Price Patrol results may have been atypical. However, personal bias may operate unconsciously, too (Tr. 2990), and other Grocery Merchandisers may have acted in a similar manner without realizing it. The important point is that the mixture of functions evident here is plainly inconsistent with the notion of impartiality that one naturally expects to be reflected in sound survey design. Accordingly, we affirm the ALJ' conclusion that Kroger s use of Grocer Merchandisers to select the Price Patrol samples did not conform to procedures that survey experts would regard as sound. (27) Representative Sampling The ALJ also faulted Kroger for failing to weight its sample results to correspond to the actual purchasing patterns of typical consumers." By the same token, he found that the Price Patrol samples generally consisted of a disproportionately large number of high-volume items and "rollers." (ID 179, 182.) The record shows that these categories of products represent a minority of the items carried on supermarket shelves. (RX 979.) Thus, it is possible that a Kroger store which discounts high-volume and roller items drastically may win" a Price Patrol survey, and yet be more expensive overall due to its higher prices on othertypes ofitems. (RX 906; Tr. 2769. While there was, again, general agreement among the expert witnesses that systematic weighting is a desirable element in survey design, the record reflects that, like random sampling, it may be impracticable' in food price 'comparisons. The prerequisite to systematic weighting is information on the frequency with which certain types of goods are purchased, Or data on the proportions of consumer spending attributable to various categories of goods. (Tr. 1053-54. outside con u!tants to wke the place of the Kroger employees presently involved in planning and evaluating the Burgoyne data. NevertheJe, we note that even if the original estimate were trebled, the slim would represent Oily a smal percentage of Kroger s 1978 advertising budget. (See ID 5 in camera.) "' Weighting is a process of statdicaUy adjusting samples-Dr data generated by samples-to- make- them rnorerepresentative of the Populations being measured. (IU 192.) In the case of a comparative foo price survey, systematic weighting requires knowledge of the frequency with which eonsumern buy particular foQ items or cat-gories of foo items. (Se Tr. 1053-54.) If it is ki10wn that fot example, consumers purchase 2 Jbs. of potatos per week, it weekly survey of thereiative cost of shopping atvariaus stores could be weightcby doubling each store s price for lib. of potatoes. There are also methods of applying data ori typical foQ purchasing patterns to the selection of the Sample iU;,ms themselves, so that the final sample wil afford comparative price data on the oVIJTall costs of shopping for foo at the stonissurvlJYCd. This type of sump!1J is caled sef,weighte. (Tr. 1053). 744 FEm RAL TRADE COMMISSION DECISIONS Opinion 98 F.

The RFPS memoranda show that the USDA has such information on food items but not on other products commonly sold in supermarkets. (RX 1003 at p. 30.) Thus, if Kroger were to have conducted a systematically weighted Price Patrol survey, it would have been required to generate considerable information on its own about purchasing patterns, not merely in Kroger stores but in its competitors' stores as well. Concededly, this would at least have been an expensive undertaking. (28) Kroger says that it attempted to weight its samples by focusing on high-volume products, and it argues that while these products may not, as a group, represent a majority of supermarket items, they do account for a sizable share of the consumer s expenditures on grocery items. Kroger also points out that its focus on high-volume products naturally resulted in a disproportionately large number of rollers on its survey lists because "roller" discounts are given primarily on fast moving goods.

Given these diffculties of producing a systematically weighted sample in a food price survey, we do not think that Kroger s Price Patrol was unsound simply because it was not systematically weighted. Whether its attempts to compensate for the lack of weighting data were adequate is a more difficult question to answer. The record does indicate that fast-moving products-the types of items that typically appeared on the survey data--o account for a sizable proportion of shopping expenditures. (CX 803.) However, the record also shows that Kroger s selection criteria were designed not merely to produce representative samples, but also to ensure that the items surveyed were those that shoppers would typically recognize. (ID 63.) In our view, the company s concern for consumer recognition has little to do with sound survey methodology, and it may be questioned whether Kroger s judgment samples might have been more effectively weighted had the company concentrated solely on developing a truly representative sample. Nevertheless, complaint counsel did not adduce evidence specifically on the adequacy of Kroger s attempts to weight its survey samples on an ad hoc basis. Consequently, we are not ready to conclude that the overrepresentation of high-volume products, including rollers, was itself indicative of unsound survey procedures. (29) Advertising Format As previously mentioned, Kroger s television and newspaper advertisements presented Price Patrol survey results in the form of box scores showing the numbers of items on which Kroger s prices were higher than, lower than, and equal to those of its competitors. THE KHUliEtt I.M.

639 Opinion The ALJ cited approvingly Dr. Kohout's testimony that these figures could not support a claim that it was cheaper to shop at Kroger than elsewhere. (!D 189.) According to Dr. Kohout, a store may have more lower-priced goods than its competitors but still be more expensive for shoppers generally if the margin of savings on the lower-priced items is exceeded by substantially higher mark-ups on other commodities. (Tr. 1052.

Of course, the issue of how to present the Price Patrol data is almost academic; since the Price Patrol surveys were not methodologically sound, no method of presenting their results could support Kroger s "proof of oVClall savings" claims. Furthermore, even if the Price Patrols had been methodologically sound, the proper question would not be whether the advertised results supported Kroger proof of overall savings" claim but whether any results tabulated for Kroger s own use supplied the necessary substantiation. The record indicates, though, that Kroger itself tabulated Price Patrol data in only one form that of the box scores showing merely the numbers of items on which Kroger prices were higher than lower than, or equal to those of its competitors. Kr. Kohout' criticisms could thus be applied to Kroger s internal records as well as to its advertising format.

We agree with Dr. Kohout that the best way of supporting an overall savings claim with survey data would be to compare the total costs of buying the survey items at each of the stores visited. Nevertheless, we are not prepared to say that the box score approach used by Kroger would have been inadequate to support a "proof of overall savings" claim. If the Price Patrol surveys had been methodologically sound, we believe that it is likely that Kroger would have had a reasonable basis to believe that the store with more lower priced items would, in the vast majority of instances, also be the cheaper place to shop.

In challenging Kroger s use of box scores, complaint counsel cite two situations where Kroger advertised it had won a Price Patrol survey but where a competitor actually had a lower total price for the survey sample. (RX 125, RX 906.) (30)Nevertheless, given the hundreds of Price Patrol surveys run by Kroger over the course of the campaign, something more than these two isolated instances is required to persuade us that the box score data could not substantiate a lower overall price claim if sound survey procedures were employed.

Summary We hold that Kroger s Price Patrol surveys were not conducted 7 - ""U n - R 2 - 48 QL ,. & Opinion 98 F.

according to sound procedures because they failed to disclose the systematic exclusion of certain categories of products to which the survey results were generalized, and because Kroger failed to ensure that those persons who selected the survey samples were not in a position to bias the results. The Price Patrol surveys, therefore, also failed to substantiate adequately Kroger s claims that it had "proof' that consumers would save overall by shopping at Kroger. As a result, the Price Patrol advertisements were deceptive and unfair. Procedural Objections Kroger has assigned error to several of the ALJ's rulings on discovery, the admissibility of evidence, and the scope of crossexamination during trial. The first of these objections concerns the ALJ' s refusal to permit Kroger to discover internal Commission documents, or portions of such, that set forth the personal views of staff members and individual Commissioners concerning the Retail Food Price Survey.'" Second, Kroger contests the ALJ's decision to exclude from the record any RFPS documents which had not been approved by the Bureau of Consumer Protection or by the Commission itself. Third, Kroger asserts that the testimony of complaint counsel' s expert witness, Dr. Kohout, should be disregarded because it was not permitted to cross-examine him on certain material that he used in forming his opinion.

We have previously affirmed the administrative judge s discovery rulings and, after reviewing the record once more, we again conclude that his actions were proper. Recommendations or other statements of opinion in internal agency memoranda, if not expllctly adopted as official agency policy are privileged from discovery absent a demonstration of compelling (31Jneed. British Oxygen Co., 83 C. 1785 (1974). Kroger does not dispute this. Instead, it argues that the Commission waived the privilege by bringing suit, and that in any event, respondent' s need for the RFPS documents outweighed the agency s interest in their continued confidentiality. As to the supposed waiver, the Commission has consistently ruled that the agency s decision to bring a law enforcement action does not vitiate any privilege that may apply to intra-governmental memoranda. Coca-Cola Co., 85 F. C. 398 (1975); Chock Full 0' Nuts, Corp. Memorandum Orde Ruling all Malians to Limit SubpuPl!w; Duces Tecltm at 2 (dated Or.L 27 , 1977); Ord..r Ruling on Responden/ s M!ition for Recansidemti,)n af the Order orOdobe 1977 and Alternatively, n"'l"esi for and Immediate Appeal (rum that Order Ht 2 (dated Jan- 9, 1978). The aclministrHt.ve law judge s orders porrnitted discovery of factual portions of internal RFl'S memoranda and of evaluativ.. portions that had been explicitly endorsedor adopted by theCommis ion as a whole Order Affirming Order Ruling all Responden/ s Molion for Productiono( lJou.ments Pursuant to *3.:16, at 1 (dated Feb. J , 1979).

THE KHOGER CO. 747 639 Opinion FTC. 747 (1973); see also Wirts v. Continental Finance Loan Co. of West End 326 F.2d 561, 563 (5th Cir. 1964). As to respondent' asserted need for RFPS information, it is sufficient to note that Kroger was in fact allowed to discover the non-privileged parts of RFPS documents, namely factual discussions in the memoranda and statements of official agency policy. It thus had access to significant portions of the RFPS materials, and we are not persuaded that it had particularized need for the privileged segments. We are also satisfied that the judge acted properly in declining to admit RFPS documents that had not been approved by either the Bureau of Consumer Protection or the Commission. Kroger argues that a number of these documents portrayed the difficulties staff encountered in developing a random sample food price survey methodology and show the staffs increasing pessimism about the likelihood of ever developing such a protocol. In our view, the rejected exhibits cited by Kroger on appeal contribute little or nothing to an understanding of the issues in this case. A number of exhibits proffered by respondent are simply preliminary versions of staff memoranda that were ultimately sent to the Commission and were, in their final form, admitted into evidence. While one might have some interest in determining how the staff refined its views of RFPS over successive drafts of a memorandum, it would be an interest purely historical in nature. We fail to see how these drafts, unapproved by the Bureau Director, shed additional light on what constitutes a sound survey methodology. We similarly fail to see the relevance of a memorandum cited by Kroger that merely advises the Commission of the staffs projected timetable for completing the RFPS effort. That the staff encountered delays in the project is already made clear from other documents in evidence. Kroger also urges that the ALJ should have accepted memoranda from the Bureau of Economics and the Office of Policy Planning, which it claims argue against continuation of the RFPS on economic and policy grounds. Neither memorandum discussed the feasibility of conducting a random sample food price surveyor the elements essential to sound survey methodology. They therefore add little to the factual development of the case. (32) The last procedural objection concerns a seven-page excerpt of a staff-authored RFPS memorandum that complaint counsel gave to Dr. Kohout to help him prepare his testimony. The excerpt was also given to respondent. 35 However, Kroger claimed that the excerpt's " The excerpt was accepted into evidence.8s RX 1022 It lists gen",ral pri.ocip!cs of survcy methodology that the Commission staff attempted to implement in conducting experimental comparative food price surveys Much of the remainder of the Ri"PS memorandum describes the practical diffculties ofapp!yinK these principles Opinion 98 F.

disclosure to Dr. Kohout had effected a waiver of the privilege for internal agency memoranda, and it accordingly sought a copy of the whole RFPS memorandum, identified as RX 1005, for purposes of cross-examining Dr. Kohout. At first, the administrative law judge denied the motion on the ground that the remaining portions of the memorandum contained nothing relevant to Dr. Kohout's crossexamination. (Tr. 1245.) Later, the judge concluded that the memorandum s privileged status had been waived, and he ordered the respondent be given the document in full. He did not permit Kroger to recall Dr. Kohout to the witness stand, however. Kroger contends that, had it been able to do so, it would have used the memorandum on cross to demonstrate that aspects of Dr. Kohout's proposed survey methodology were not viable. Respondent argues that the denial of this opportunity for effective cross-examination requires reversal. We are inclined to agree with Kroger that complaint counsel impliedly waived the privilege for the full RFPS memorandum and that the AU should have permitted Kroger to call Dr. Kohout back to the stand for additional cross-examination. We need not pause to resolve these issues, however, for it is clear to us that whatever errors may have occurred did not substantially prejudice Kroger s rights. Kroger cross-examined Dr. Kohout at length concerning his opinions, and though it was undeniably deprived of the impact of confronting the witness with the staffs actual experiences in conducting the RFPS, it was at least able to question him on some of the problems in his recommended methodology that were highlighted in the staff memorandum. More importantly, during defense hearings, Kroger s own witnesses were able to state their criticisms of Dr. Kohout' s methodology, and Kroger introduced a survey that it had conducted in conformity with Dr. Kohout's recommendations in order to show the anomalous results that they could produce. (33) Disqualification Kroger argues that AU Hyun should have been disqualified from participating in this case because he served as an attorney-advisor to former Chairman Engman from March 12 to September 30, 1973 during which time he would have had access to staff and Commission memoranda concerning early stages of the RFPS. Kroger contends that staff predictions about the viability of the RFPS were uniformly optimistic during this period and that Judge Hyun s exposure to ,n (f a waiver occurred, it was not imply because the excerpt was disclosed to cumplaint counsel's witness; rather it was because the witness also relied to SOme extent on the excerptin forming his opiniun and then gave test.imony On t.hClt opinion. Se Uniled Stalesv. Alvarez 519 F.2d 1036, 104546 (;!d Cir. 1975); La/anre Gn;sjeun Mfg C(J. v. HnhermanMfg. ' 563(S. NY. 1898) . . .

THE KROGER CO. 749 639 Opinion such optimism preconditioned him to believe that it would be feasible to conduct a random sample comparative food price survey. Alternatively, Kroger argues that it should at least be granted additional discovery of Commission files so that it can more precisely show, as a predicate for its disqualification motion, that the RFPS and this case are factually related.

Standards for the disqualification of an ALJ were recently considered in GraZier v. FTC, 615 F.2d 1215 (9th Cir. 1980). At issue there was the construction of the third sentence of Section 554(d) the Administrative Procedure Act:

An employee or agent engaged in performance of investigative or prosecuting functions for an agency in a case may not, in that or a factually related case participate or advise in the decision, recommended decision, or agency review pursuant to section 557 of this title, except as witness or counsel in public proceedings. The court concluded that this provision, like the rest of Section 554(d), was designed to ensure that cases were not determined either by persons who had developed a wil to win through previous partisan involvement or by persons who would be likely to interpolate ex parte information into the determinative process. (34) The court, therefore, held that Section 554(d) precludes former attorney-advisors to the Commissioners "from participating (as ALJs) in the adjudication of cases in which they have actually performed ('investigative and prosecuting ) functions, and in 'factually related' cases. " 615 F.2d at 1221. The court further said: Once an attorney-advisor is shown to have been ' engaged in the performance of investigative or prosecuting functions' through prior acquaintance with ex parte information, 554(d) says he ' may not. participate or advise in the decision the case. Id.

In such circumstances, disqualification is required even if the judge cannot recall the ex parte facts to which he or she was exposed. The court did conclude, however, that the party moving for disqualification bore the burden of proving that the ALJ actually was apprised of disqualifying information; one could not presume, for example, that the former attorney-advisor had been apprised of certain information simply because he or she would have had access to it as an advisor. To aid the movant in establishing actual involvement GraZier held that the agency involved was responsible for producing sufficient evidence to permit an accurate determination under Section 554(d). It suggested that an agency could fulfil this duty by responding in the form of affidavits as to the existence and extent of the ALJ's prior involvement with the litigation at hand or a factually related case.

750 FEDERAL TRADE COMMISSION DB CISIONS Opinion 98 FT.

Prior to the Ninth Circuit's decision in GraZier the Commission considered and rejected Kroger s motions for the disqualification of ALJ Hyun and for discovery of Commission files on the Retail Food Price Survey. 93 F. C. 202 (1979); 93 F. C. 302 (1979). After Grolier was announced, however, we reconsidered our ruling and directed complaint counsel, ALJ Hyun, and the Commission Secretary to file affidavits on the factual relatedness of the Retail Food Price Survey to the Kroger case and on the likelihood that Judge Hyun would have been apprised of information concerning the RFPS or Kroger pricing policies while he served as an attorney-advisor to former Chairman Engman. Order of June 1980. Complaint counsel's affidavit denied that there was any factual relation between the RFPS and this case, and asserted that the origin of the Kroger investigation was independent (35Jof the RFPS. ALJ Hyun stated in his affdavit that he had no documents in his files revealing any prior involvement on his part with the RFPS prior to being assigned to this case. The Secretary s affidavit similarly averred that no Commission records disclosed whether Judge Hyun had participated in the RFPS. The Secretary did state however, that his inquiry revealed that 55 documents concerning the RFPS or Kroger s pricing policies had been sent to former Chairman Engman s office prior to and during ALJ Hyun s tenure as attorneyadvisor.

On February 24, 1981, we directed the Secretary to forward these documents to the Commission, and on April 3, 1981, we ordered that they be made available to the parties in camera for their comment. The parties fied their briefs on May 8, 1981. Kroger asserts that during Judge Hyun s tenure as attorneyadvisor he was likely to have been exposed to the RFPS memoranda that the Secretary has forwarded. Kroger also asserts that these parte communications were disqualifying because there is sufficient evidence to show that the RFPS is factually related to this case. In this regard, Kroger notes particularly that staff attorneys in the RFPS project collected information specifically about Kroger " In our previous ruling, we heldthatKTOger had neither due process nor s1",tutory rights to discovery Or to the disqua!ifir.ation of ALJ Hyun. Our reconsideration of that ruling was limited to the construction of S!;di()o fifi4 of the APA. Our previous holding on due process right,, to discovery or disqualification here is left undist.urbed ,. ALJ Hyun had previously st3ted that he had no noeollection of being exposed to tho' RFPS during his t!nure as attorney-advisor. s.",Order Ruling (In Respondenl s M"ti,Jf (or Produdi,,, Purslwnl 10 Sedirm 36 and Certi(ying the Rulings10 the C()mm siol! (Jan. 15, 1979). '" Kroger contends that our decision to release these documents, which include staff memoranda and Commissioner circulations about the RFPS, demonstrates t.hat. t.he ALJ erred in denying Kroger s pm-trial di;;covery requ!;sLs for such materials. We have already affrmed the AlAI's rulings in this opinion. See text acr:mpanying nats 3:1- supm. Our decision to release certain RFPS m!;moranda to the parties;; for the limited purpose of resolving Kroger s disqualification motion is not inconsistent with thi;; affrmance. The release we hav!; ordered was intended solely to iluminate any ex parle knowledge that Judr:e Hyun may have had concerning issues of fact in this cas THE KROGER CO. 751 639 Opinion discount pricing policies and practices. Kroger also asserts that discussions of survey design in RFPS memoranda were directly relevant to issues in this case. (36) It is unnecessary to dwell on Kroger s arguments that ALJ Hyun was likely to have seen the RFPS memoranda that were before the 0 Disqualifi-Commission during his tenure as an attorney-advisor. cation would not be warranted even if Hyun actually had read all of those memoranda.

For one thing, we do not believe that the RFPS qualifies as a factually related "case " which is a prerequisite to a successful disqualification motion under Section 554(d). The RFPS is certainly not the type of adversary proceeding that the word "case" typically connotes. And neither the language nor authoritative interpretation of Section 554 suggests that "case" should be given an expansive reading here. By its terms, Section 554 applies only to adjudicative matters that must be decided on the record after opportunity for an , the Attorney General'agency hearing. 5 U.sC. 554(a). Furthermore Manual on the Administrative Procedure Act advises that "(tjhe limitation of the prohibition against consultation to those who performed investigative or prosecuting functions ' in that or a Id. at 54 n.factually related case,' should be construed literally. (1947).

However, even if we assumed that the RFPS was a case, we could not conclude that the other conditions to disqualification under Section 554(d) have been fulfilled here. Under Grolier Judge Hyun can be disqualified only if it is shown that (1) the RFPS is a factually related" case, and that (2) Judge Hyun was "engaged in the performance of investigative or prosecuting functions through acquaintance with ex parte information" concerning the RFPS. (37) In our June 5, 1980, order, we stated that cases may be factually related if they arise from a common nucleus of operative facts; and, assuming that the RFPS was a case, we suggested that it might .0 Respondent's argument that Judc;e Hyun received ex parte information on th" RFPS is based largely on a letter dated Auc;ust 4, 1980, U) Kroger col1nsel from former Chairman EngmaI1. Se" Exhibit A t.nR".'I"Jlse "r 1/", Kmg r C(Jmpany 10 Arfidav;ls Filed Pursuan.tto the Comm;s. i",,: J,,,,, 1980, Urder (fled Al1g, 8, 19HO). Engman stated in this lett"r that he has rid record ofMr. Hyun s working on Rr'PS mat.t.ers as on '1UorJwy-advisor However, Mr. Engman also stat.ed that he discLlr;ed "significant rniltters coming before the Commission at. confercnt:es with all Ihj J attorney-advi ()rs" and "rtJhus it i possible that lRFI'S memoranda "I' Mllrch 22 "nd April 13, 1973J wlluld h"ve been di cussed "I, such II confer"""e. " Kroger asserts t.hat the RFPS Wa5 it)d ed " ignificant. matt.,r for the Commission during early 19n, so that ALJ Hyun must have be,,,, acquainted with I. RFPS while he worked in the Chairman s office B:ven if Judge Hyun did learn of the RFPS through the former Chairman s staff meeting, the extent or his knowledg-e is sti1J entirely unknown. Mr- Engman s letter do,'s not indicate whether ti", supp"sed discussion or the Rr' PS at these staff m..etinJ;s wos lihly to have been "ursory or ddailed- The fact thatAL.l HyuCl cannot remember leilrning anything about the HFPS during his !.nure ae; il!.orney-advisor sug-gcsts thatany slaff discussiof1 of the project was !ess than extetlsive " See also Gralier Docket Nt). H879 Order Denying M"I;(", /0 Disqltaliry ,lwi.gc vol! Brand at 10 12 lfiled Au!; 19B1) 752 DERAL TRADE COMMISSION DECISIONS Opinion 98 F.

be factually related to the present proceeding if it involved Kroger pricing policies and practices.

But, beyond showing that the RFPS was factually related to this case, Kroger must also show that Judge Hyun was performing an investigative or prosecuting function through his exposure to parte information concerning the RFPS. Common sense suggests that not just any ex parte communication concerning the RFPS should suffice to disqualify Judge Hyun. Rather, it appears that the necessary ex parte information must concern in some way those operative facts that are common to the RFPS and this case. Otherwise, there would be little sense in conditioning disqualifieation on proof that the ALJ had previously investigated or prosecuted a factually related case.

We also believe that disqualifying ex parte information must concern the facts in issue in the instant proceeding. This standard is embodied in Section 554(d)(1), which forbids an ALJ from engaging in ex parte consultations with any party or person on a "fact in issue." It is an appropriate standard to apply to this case for two reasons. First, it would be anomalous if an ALJ could be disqualified for having been exposed in a prior case to ex parte information that was more innocuous than that proscribed in Section 554(d)(1). Second, the GraZier court itself relied on Section 554(d)(1) in defining the scope of an attorney-advisor s "investigative or prosecuting functions." 615 F.2d at 1219-20.

Thus, even if it is assumed that the RFPS is a case, ALJ Hyun can be disqualified only if Kroger shows that the RFPS is factually related to this proceeding, and that Judge Hyun s supposed contact with RFPS memoranda during his tenure as attorney-advisor exposed him to facts at issue here.

Kroger attempts to demonstrate factual relatedness by drawing our attention to a ,January 29, 1973, memorandum from the Bureau of Consumer Protection, advising the Commission whether the staff had investigated the possible deceptiveness of low price claims made by a certain grocery chain. The Bureau s memorandum stated that the staff was not investigating that particular company s claims but was instead developing (38)the RFPS protocol to secure information about which supermarket chains in given metropolitan areas actually do provide the lowest prices. The memorandum noted in passing that "many other firms, including Kroger, Safeway, Stop & Shop and National" were making discount claims. No other mention is made of Kroger in the memorandum. We cannot conceive how this document could show that the RFPS and the present case share a common nucleus of operative fact, much Jess the specific facts in .

THt; KH.UUt;K CU, (b;J 639 Opinion issue. A simple reference to the existence of Kroger s low price claims in a memorandum describing the general purpose of the RFPS is plainly insufficient to prove factual relatedness. Kroger has also pointed out that, in the latter half of 1973, RFPS staff obtained data directly from Kroger about aspects of its pricing and advertising policies. This information might well establish that the RFPS and Kroger are factually related in the sense that somewhere in the RFPS files there is material that may concern operative facts in this case. However, for the purposes of the instant motion, the critical question is whether Judge Hyun was apprised parte of this information about Kroger s pricing and advertising. There is no evidence that information was ever conveyed to the Commission, so it is impossible to say that Judge Hyun learned of it when he was an attorney-advisor. Indeed, Kroger does not even argue that the ALJ was apprised of the information ex parte. Respondent does argue, however, that in 1973 ALJ Hyun was likely to have been exposed ex parte to RFPS memoranda concerning proper survey design and the feasibil ity of developing a statistically valid food price survey. Respondent contends that such memoranda concern the issues raised by this case and that the ALJ' s involvement with the RFPS would have led him to believe that a reliable low-cost food price survey could be accomplished with ease. However none of those memoranda contain data that would have related the RFPS to facts in issue here. The memoranda do not, for example discuss Kroger s pricing practices or its advertising campaigns- Rather, they concern only general issues of survey design. These matters are certainly relevant to the appropriate level of substantia tion for survey claims. However, they do not concern the factual issues that were litigated here, such as the meaning of Kroger advertisements and the particular procedures Kroger used to compare its food prices with those of its competitors. Accordingly, the RFPS memoranda that were before the Commission when ALJ Hyun was an attorney-advisor did not contain information that would justify his disqualification under Section 554(d). (39) Even if it were assumed that these memoranda concerned facts in issue here, we believe that Judge Hyun s disqualification would be unnecessary. As stated in Grolier Section 554(d) is designed to ensure that decisionmakers have neither the will to win nor the ability to interpolate facts not already in the record. Kroger has not .'- Se Attorney General's Manual on the Administrative Procedure Ad 130 (1947) (r!' printing Appendix to Attorney General's Statement Regarding R.wised Committee Print of October 5 , 1945) "The term ' fact in issue' is used in itstechnica!, litigious senf;." This suggests that Section 554tdXl) prohibitsexpa rleconsultationsonJyasto adjudicative" facts, that is "facts concerning the parties-who did what, when, how, why, with what motive or intel1t 2 K. Davis, Administrative Law Treatise Section 12:3 (1979) Opinion 98 F.

alleged that ALJ Hyun had a will to win. As to the likelihood that facts not on the record would be interpolated, we note that the most comprehensive RFPS memoranda produced by staff have actually been admitted into evidence. RX 1000 , 1003. See Grolier Docket No. 8879 , Order Denying Motion to Disqualify Judge van Brand at 9 (filed Aug. 12, 1981). Thus to the extent Judge Hyun s possible knowledge of the RFPS in 1973 matched the information in the adjudicatory record here, there is no risk that Judge Hyun would have been able to "interpolate facts and information discovered by (him) ex parte and not adduced at the hearing. . . . Grolier v. FTC, 615 F.2d at 1219.

We therefore conclude, on the basis of the affidavits submitted and the RFPS memoranda forwarded by the Secretary, that the disqualification of Judge Hyun is unwarranted. Kroger, however, has raised objections to the adequacy of the Secretary s search for documents ordered by the Commission in its orders of June 5, 1980 and February 26, 1981. Kroger complains that a search should have been made of the files of the Division of Management and the Assistant Director for Legal Coordination, since both offices participated in the RFPS. Neither office currently exists; their files have been merged with those of the Executive Director, whose records were searched. Kroger also contends that the files of individual Commissioners offices should be searched. Not wishing to leave any stone unturned we have asked the Secretary to review each Commissioner s files for additional documents concerning the RFPS or Kroger pricing policies that would have been before the Commission during ALJ Hyun s tenure as an attorney-advisor. No new documents were found. (40) Finally, Kroger makes two requests in aid of its effort to establish the factual relatedness of the RFPS to this case. First, it has moved that we place on the record a staff memorandum of October 31, 1978 discussing a request by an attorney for clearance to participate in this proceeding. Second, Kroger asks that, if its disqualification motion is denied, it be allowed discovery of all the Commission ,,, In this regard, it is ..Iso wurth noting that much of the RFPS stafrs thinking about survey design and the feasibility of statist.ically v..Jid f,)od price surveys were actually matters of public record at the time ALJ lIyun would have sen them. Se the C_ommission s proposed prutocol for the RFPS which was published for comment at 38FR 1543 (Jan. 151973).

.. It is alsodiffcult to conceive how Judge Hyun s supposed ex parte involvement with RFPS information actually injured the respondent. While ALJ Hyun concluded that statistically sound comparative foo price Sllrvl'YS are feasible, his initial decision ccrtainly discloses ao appreciation of the practical problems they present. Moreover, the ALJ's recummended order does not require Kroger to adopt the random sample methodology on which the RFPS was based. We t.therefore fail to SHe any injury from the allegedprLrtf!!'communication. Where it can be shown that an ex parte contact has in fact resulted in no substantial prejudice to a party, there can be no revl'rsib!e error- 5 US- G 706; Marathon Oil Co. v. EPA, 564 F.2d at 1265 (9th Cir. HJ75); see also Attomey General's Manual on the Administrative Procedure Act7:J-74 (1947) THE KROGER CO. 755 639 Opinion RFPS files so that it can better demonstrate the factual relationship between the RFPS and this case. These requests are denied in view of our conclusion that, even if the RFPS does constitute a "factually related case " none of the documents Judge Hyun might have seen as an attorney-advisor contained disqualifying information. (41) Order Provisions The ALJ issued a cease and desist order that would impose detailed and comprehensive restrictions on Kroger s use of comparative food claims in future advertising. The order extends to any claim that Kroger s prices are lower than those of particular competitors or of the competition as a whole in a particular area. The ALJ' order would prohibit such claims unless: (1) they are based on a survey conducted according to sound survey procedures; (2) the claim presents the survey s results fairly and impartially; and (3) the survey s results and methodology are available for inspection at Kroger stores.

The ALJ defines sound survey procedure somewhat differently depending on whether the comparative claim refers to survey results or not. Where the comparative price claim does make such a reference, the underlying survey must conform to specific methodological criteria concerning issues such as the selection of survey samples and the competitors to be compared. If the survey fails to meet these criteria, Kroger may not refer to the survey results in its advertising.

Comparative claims that do not refer to survey data must still be founded on substantiating survey results. However, the sound survey procedure requirements are slightly more relaxed for such advertisements. If the supporting survey does not meet each of the methodological criteria referred to above, the general comparative price claim may be advertised if it also discloses each criterion on which the survey is deficient.

The order would prohibit the dissemination of any comparative retail food price claim-whether or not it refers to surveys-without an accompanying disclosure that the amount consumers will save at Kroger depends on what they purchase. This and any other disclosure required under the order would have to satisfy detailed guidelines on acceptable format in print and broadcast advertising. Kroger has objected strenuously to the ALJ's order, arguing that Opinion 98 F.

its only effect will be to eliminate comparative retail food price advertising." Respondent's first objection to the order concerns its breadth: it is not limited to covering price claims based on survey sample data, but would apply to any claim that Kroger s prices are lower or lowest, presumably (42)even when the claim concerns only a comparison of prices charged by Kroger and one competitor on a particular food item. Kroger argues that such simple one-to-one price comparisons hardly warrant the elaborate sampling procedures and cautionary disclosures set forth in the ALI' s order. Kroger further contends that the sampling procedures are unnecessary-and linwise-ven in the case of survey-based claims. The order provides that samples should be drawn on the basis of either random selection or the judgment of an independent consultant. As noted above, respondent argues that these alternatives are too expensive to be employed and that it can conduct comparative food price surveys only if it can use its own employees to select judgment samples. Of course, the ALI's order would allow Kroger to select its own judgment samples provided that its price claims did not purport to be based on survey data and its advertisements disclosed that the underlying survey samples were not randomly selected or chosen by an independent party. Kroger argues that this whole disclosure scheme is unrealistic, however, since the number of disclosures that the order would require for an affordable survey would be excessive. Such disclosures would, moreover, have to be presented in accordance with guidelines specified in the order; for example, print disclosures would have to appear in 16-point type and TV disclosures would have to remain on the screen for the entirety of the commercial. 40 (43) Complaint counsel, in their answering brief, have proposed a number of changes to simplify the ALI' s order." The Council on ., Kroger s concerns about th., ALJ's order are &hoed in an t.micus brief fied by the Foo Markp.tjng Institute.

00 Kroger also contends t.hat the AlAI's order violates the First Amendment cause it is not the least restrictive method of regulating the company', advertising. CentralIn Hudson Gas and Electric Corp. Public Service C"mm ,ion S. (1980), the Supreme Court held that the First Amendment protects truthful commercial speech from govlornmental regulation that is not. essential to the promotion ofa substantial governmental interest. However, the Court also held that nO constitutional protections extend to speech that is dll:eptivc or related to illegal activity. It lies within the C"mmission s power to prohibit such speech and to frame orders that are effective in preventing the recurrence of misleading advertisements.Jay Norris, Inc. . F"' , 598 2d 1244, J251 "i:-I (2d Cir 1979). We find it unnecessary, however, 1. resolve these constitutional issues, since the order we issulo toay is a significantly narrowed version of the ALJ' 8 proposal, and does rell&t the least restrictive alternative to regulating survey-based food price elaims. The modifications Wlo havlo mad" were impelled not so much by constitutional constraints as by an apprloeiation of the nee to allow for flexibility in thlo design of comparative foo price surveys. ., Among them is thlo ddction of the mandatory disclosure that the amount consumers would s."1ve by shopping at Kroger would depend on what they purchased. Complaint counsel felt that this disclosure is nol essential sim:e iL message was fairly apparent. Kroger aTb"Uesthat this "concession " on complaint counsel's part undermines the del.rmination that the Price Patrol advertisements reprp.Slonted Consumers would save overall shopping at Kroger. (RRB 16-19.) We do not agree. The belief that. one will saVio uvp.rall at Kroger is not at all inconsistent with the logical notion that the amount one saves wil depend on what. or how much onlo buys l"t1 .tt(Utj jt LV. will 639 Opinion amicus briefWage and Price Stability (COWPS) also submitted an that supports those suggestions but recommends additional modifications to ensure that truthful comparative food price advertising is not discouraged in the future. COWPS has suggested that the order not attempt to specify exactly how Kroger must conduct a retail food price survey, but attempt instead to simply guide it to a number of acceptable alternatives for such surveys. As was stated previously in this opinion, the Commission is fully cognizant of the importance of avoiding unnecessary restraints on truthful, comparative advertising. It has therefore endeavored to ensure that the order issued in this case is no broader than necessary to address the violations that have been established in this case or closely related conduct. The Commission s order is thus much narrower and simpler than that issued by the AL.L (44) Paragraph I of the Commission s order sets forth the definition of survey-based food price comparison claim " which in turn defines the breadth of the order as a whole. While the ALJ's order would have applied to any comparative food price claim, the definition used here provides for a more limited reach. The definition covers only comparative price claims that both refer to survey results and project the results obtained from a specific survey sample to items not included in the survey. As limited, the definition is focused precisely on the distinctive characteristics of Kroger s Price Patrol campaign. The definition would not cover a variety of comparative price claims commonly made by supermarkets, including Kroger that do not refer to price surveys. For example, the definition does not apply to 'trifle shot" claims-simple comparisons made of an advertiser s and a competitor s prices on a particular food item or product (e. a comparison of prices charged by two stores for a 5 lb. roast). Nor does the definition cover comparisons made on an item by item basis of the prices charged for a number of listed products, as often appears in newspaper food sections. Finally, it does not cover claims, unaided by reference to survey data, that the advertiser is e. RRB, Appendix B.)lower in price than its competitors. (See, Such general comparative price claims must, of course, be reasonably based, and it may weU be that adequate substantiation must consist of a survey of some sort. However, this case does not present such simple comparison price claims, and we are not prepared to declare in the abstract what substantiation those types of claims would require. They are thus wholly outside the scope of the order we have issued.

Part II of the Commission s order contains aU of the substantive provisions affecting Kroger s future dissemination of survey-based Opinion 98 F.

food price comparisons with specific competitors or with the competition in a particular locale. It begins by prohibiting such claims unless they are based on a survey, test, or comparison that is reliable and competent. This general admonition simply reflects established law that tests and surveys necessary to substantiate a claim be designed and implemented in a trustworthy manner. Litton Industries, Inc. Docket No. 9123 (Jan. 5, 1981) (97 F. C. 11, petition for review docketed No. 81-7148, 9th Cir., Mar. 12, 1981. This provision is also comparable to the ALJ's requirement that surveys be conducted according to sound survey procedures. However, in contrast to the ALJ, we have not attempted to spell out the specific procedures that are essential to a reliable and competent food price survey. As we stated above, (45Jsound methodology-at least in the context of food price comparisons--epends partly on the practical obstacles presented in a particular case and partly on the advertising claims ultimately disseminated about the survey. These factors are infinitely variable, and no single list of essential" survey techniques would be suited to each possible circumstance. We have, however, specified procedures to be followed in selecting survey samples, since that is the area in which Kroger s Price Patrol design was deficient. The ALJ's order would permit Kroger to assemble its own samples if it uses random selection techniques, but would allow only an outside consultant to develop judgment samples. Kroger has protested against this restriction on the ground that random sampling is infeasible and the cost of training a consultant to pick judgment samples exorbitant. The record clearly indicates that significant obstacles stand in the way of random selection least in the context of retail food price comparisons. The expense of training a consultant is harder to gauge, but complaint counsel appear to concede that no consultant is currently prepared to select its own judgment samples. Finally, the deficiency we found in Kroger s methods of selecting Price Patrol samples was not actually the company s reliance on its employees' expertise to make selections but the fact that they had responsibility over both selection and pricing. Accordingly, our order permits Kroger employees to continue to select representative judgment samples. However, the order also provides that the Kroger employees responsible for pricing merchandise must not know which items have been selected for '" Conse'luenUy, we have not required Krogp.r to follow criteria specifi", by the ALJ for selecting the competitors to be compared in surveys, or for ideatifyiag the iadividuaj stores to be surveyed- Our omission ofthes. requirements does not mean that we believe they need never be met in foo price surveys. Rather, we fed that their significance may vary from casp. to case, and we hesitate to require that they be observed in every instance. l"urthermore, thp.re is nO p',,-uasive evidence in the r=ord that Kroger excluded significant competitors from its surveys or that its Price Palroll",n; were dir..ted to competitors' stores that ch,ng..d atypically high prices THE KROGER CO. 759 639 Opinion survey samples until after the survey is completed. This requirement is needed to ensure that the prices on sample items are not cut disproportionately as a means of enhancing Kroger s performance on the survey. (46) The Commission order also provides that whenever judgment samples are used, the items whose prices are compared must be identical or substantially similar, or dropped from the survey. This provision appeared in similar form in the ALJ's order, and is, we think, clearly warranted in connection with the use of judgment samples. We do not understand Kroger to object to this provision in principle. Kroger has complained about the order s definition of substantially similar " which it contends will favor the cheaper grocer with lower quality goods. This possibility certainly exists. However, it should be remembered that the surveys under discussion here are designed to compare price, and if it is impossible to hold quality constant among the various stores being surveyed, it seems reasonable to ask that the company conducting the survey not favor itself by comparing the prices on its cheapest brands with those of the competitor s highest quality brands.'" On the other hand, it is quite understandable that Kroger would not wish to conduct surveys which benefit merchants with lower quality goods. Kroger s alternatives under the order are to exclude such merchants from its surveys or to exclude specific product categories in which clear quality differences appear. In the latter case, as is explained below, Kroger must limit its survey claim so that survey results are not projected to the products that have been excluded.

The requirement that identical or substantially similar products be compared does not extend under our order to surveys based on randomly selected samples. The record indicates that it is extremely difficult to develop a randomly-drawn sample that is both generally representative of one store s product pricing policy and consists of items that can be found on the shelves of other stores. One apparent solution to this dilemma recommended by complaint counsel' s expert witness is to abandon any effort to compare the prices charged by different stores on the same goods, but to select equivalent though not identical random samples from each store. ('fr. 1074-1084. While it would thus be impossible to compare stores' prices on an 00 The ALJ's order provided that when an identical or substantially similar item could not be found at. a particular store, the it.m would have to be drops"d from tho survey complet.ly and it.s deletion noted (presumably in the survey descriptions to be poswd in Kroger stores rather t.han in the advertisements themselves). We hllve not incorporawd this provision iol." our order. It may well be that respond..nt s inability to compare the prices on all sample it.em!; at each competitor wil not seriously affect the reliability of iL !;urvey results - -the fact t.hal. 000 or t.wo items of a large sample may be unavailable for comparison at one store does 001. dearly suggest that the samples compared in oth..r stores must also be mudified- Huwtover, a high incidence or " prolonged period of unavailability may require adjustmtonLs in the survey sample. Opinion 98 F.

item by item basis, the stores' general pricing policies could be compared. The order accordingly allows Kroger to conduct such a random sample survey, provided, of course, that it is reliable and competent. (47) The second paragraph in Part II prohibits the dissemination of survey-based food price comparisons that do not fairly present the results of the surveys on which they are based. It also forbids claims that generalize survey results to product categories that were systematically excluded from the supporting survey. This specific restriction responds, of course, to Kroger s practice of generalizing Price Patrol results to all of its food products, even though meat produce, and certain other items were customarily kept off of the 50 Kroger can ensure that its survey-based foodsurvey sample lists. price comparisons are not generalized to products that were systematically excluded either by limiting the advertisement' s affirmative claim to begin with (e. Our survey shows we have the cheapest canned goods in town ) or by prominently disclosing the excluded product categories.

While we have not required Kroger to make such disclosures in any particular format, the order does state that they must be clear and conspicuous. To be clear, the disclosure should fairly and comprehensively communicate what goods have been systematically excluded. Categorical descriptions should suffice in most instances. For example, in the Price Patrol campaign, it would have been adequate to have the advertisements state conspicuously that meats, produce, and certain dairy products have been excluded from the Price Patrol surveys."51 It is more difficult to prescribe what steps must be followed to ensure that a disclosure is conspicu ous. That determination can be made conclusively only after assessing the net impression of a given advertisement. However, in print advertisements, the size and color of the disclosure relative to the rest of the copy and the disclosure s proximity to the surveybased claim will obviously be important factors. They wil likewise be important in broadcast advertising although in radio and television commercials additional variables-such as when the disclosure occurs in the advertisement and whether it is announced over sound effects-will also playa role in determining the disclosure s conspicuousness. (48) '" The limitation On generalized claims applies unly tu tht, extent that categories of produ"L hav" be"n exeluded systematically- We recognize that no sample or serit'5 of amplp.s is likely to include every produd stocked in" grocery store, and we do not me"n to suggest that that fael alone forbjd a grocer from projecting the results of an otherwist' valid survey to his store as a whul., " Some of Kroger s post-1976 newspapt'r adverti rnent'; contait1ed clearly worded st.at!;m"nL uf whicb categories of items had been excluded. Ilowever, those rJisdosurt'5 were neither conspicuous nor regubrly made 1m; KROGER CO. 761 639 Opinion Kroger has voiced no objection to the provisions comprising the remainder of the order. These sections are fairly standard and are designed to ensure respondent' s compliance with the order s substantive provisions.

Lastly, when this opinion was in the final stages of consideration by the Commission, respondent submitted a request for permission to file a brief addressing the question whether, in view of the decision in Francis Ford, Inc. v. FTC 1981-2 Trade Case 64271 (9th Cir. 1981), petition for rehearing pending, Nos. 79 76-7, 79-7654 (filed Sept. 22, 1981), certain issues raised by this case should properly be addressed through rulemaking. The Commission believes that Francis Ford, Inc. v. FTC, supra does not correctly state the law and, in any event, the instant case is clearly distinguishable. Moreover respondent has not offered sufficient justification for a further round of briefing at this late date. For these reasons, respondent's request to file an additional brief is denied.

" One aspet of the Al.J's order that was designed in part to promok compliance was a requirement that Krw er post detailed descriptions of its survey methodology and results in its st.ores. We have decided, however, not to include this pruvision in our final ordef So long- as Kroger s surveys are competent and reliable .md their results fairly presenk.., as the Commissioo order requires, thine is 00 !cgll! fwed for Kroger to tell consumers exactly how its surveyfi are designed and precisely how the sample data break down. The disclosure of such informlltiOIl would be salutary, of course, but it does not appear necessary to prevent deception. 35 7- 1+44 0 - 82 - 49 QL 3 . .\,,: .\, j_., ....,. ....,._.... .. :. .. ..._ Opinion 98 F.T.

Appendix Selected Price Patrol newspaper advertisements and television commercial scri pts.

W(lITU R(NJo ! tcu Coj P?:! :;:!.l lnc. .I ., 18J-660 \.n"'T" A"'''' 8H C\W!T ;;ron.rISou,hlnnd ! - rL Q.......,.""_"""""","-,,OO tad! rv- 30-.s01I- UiC6rn :10 D-,,,.,, _.:_'''''''''JOJ. .l''' . mu Gu..tPr.1c"P'troll.. O.-.--..-..,(U ,.."... Irl.-/IIT Athn,.-- 'SBI1I!1GIVWXIAICC: M....."""'T"u- UATt.'W! 6/20/l6- /76 'n"g't.110 c'" .o11U.ID//.TAOEOCGI' O LM ua! o \' VISUAL AUDIO 1.., oro; a. MCU (\f GUESt 1. GUESt, Belh.... ",,,... the !(ros"'" Price P"trvl I .. r..lly fu.d",..""'hohBl th.l"" d"u In that. ' I" . 1 'Z. PU mrr fa R.'1EA E!\ 2. SANDY: 111, I' ", Sandy of the It.-OS''' Pdca Pan" OOST1B!:IOESAIrY. i' . LOGO I'A"GS !.O"', LIMJ0 Ind; thh 11 l1.-a.hylor.' Wa a hd h.,. .h. SHEE: . M1.OI/ I LlRG!: Bw..UP GP OIA!U .1o!:!.I,,,,a,,..klyp..leach.ck. .,L\ ' . CIJT m Ir-' . TAYLO \-1m J. GU!:T; 1 h d them Ick ..i"u en ISIJ H "'a ." .. Q!,\U!H:.3ID . OH l.IM AI'O R.GUl f'LL'i at nai at thl ""1e.. A .l..ntb tee" DJ!... " SHE Qnl"s m Llm!:!I COLlh ..Hlil.n and K""l:"" ..."11..didhoVQ m"... 1"" 1l""ich thin Iny "f tho "th...Itc"., w. tle,,..d '+. . CUT BACt -:0 SHG1 Dr 2 SANOY: S.athacopyo!ourlate.tr.I'",.t \/CMEN OJARt&LOGD at Kro\Io.. .nd ..0 Io11y Ie many h""",". r" like 11..1. Tlylor b li.v.. Kro b" nlil.. .! Atlant. 1"" I'dc. l..d...:

eC/tk O/l(,l1e 91f1Z""'' D..!(J KROGER CC!FAS'f I Io- /'z ,3 7 / p- /6 ; ;1 40' l...""

1'- 101-' ,,,:,..!(:;:....,..,... ..;;;.....,\ . ,j .\ ,;.. :::.:::::... .. . ,. .._::, .. ....,,.....,... .. ... : , . THE KROGER CO. 763 639 Opinion ''IU,,, t111I111.ntn.o . lnc. ..1 . \ tG C," Cii,'npbt!Il-JlUhun. iA1;r:.. ' .."un"!.,""

i') lor,.. T' \\u"n ".,:u, JOItlll haBori$outhhnd?B1-6SB' Ours.- r...)Jln- , tool TV-JO-50U-IOl prico p"trol 11 O_r- _N.nU.- Y'- 11I. cr "tl.nt&.-HSCIWUI't1 ""-U-"" . C. .. ",....'T"'''' OAlITlM!4/lS/H-S/1/16 ;- 0 ' Qco ten OI.I1Wld//.TIJDtDn OtM , Oro/SUG( J..". VISUAL AUDIO l.. . QVF.. ON FULL IIQT OF 1:. Th.. b...t "'''y to find out ",ho r".U:, h.. L:IES IIITII U)Q th.. lo:"p..tce. 1,.th.. Mt&Ju t ': . - .r.."i.to"h.olt th.. l...dln') food ' I 'Pr- ..t.ihr. ..."h ......k. 1. DIR.croa' s CUOIC!! 2. Th..t' "'h.t:tJ..llroq..rpri"lip"trold",,". 1. COiP"R150IZCU""1' W.. "",, lud pol"". on l O I"' \'l"" na.. SUP-CII. . PRICESCIICCMJ:D r:Item1iO. APIlU. 191G , br.ndlt"",.. ond, found lira,,"" low." on tilt. 'n.o,.. ",any it..",..,. . high"'" 0.,. t i.., "".,y ItarM.. v." i.,: and to.. ."m" On this ",,,ny i t:I "Iola CHA=' 0,-,,.. """in. IIro')"" "'". low"r on moro qroc..oy '1. It,,,,.th,,nanyoth,,r.tor...,,ch,,ck.d.. S..."'hy t.c 5 CIIOIC!: a".d our o:ompl"t.. r"port at IIroq3r, .." b"U..v.. Kroq.:. i. olnq ..hat tt taka. tc atayth. Athnta 10" pd."" laad,,".

co,1l/7I7S CDh'1t"J.nf! .r r ", KROGF.R CO).UN n.. /"'t /!.537/f1. il \I!\ IIH 40827 ::g:;.... ., Opinion 98 F.

.;::I ' Jo:;

=coi''" - - I...

=:;:c j THE KROGER CO.

Opinion -is ;' O :_, :. ::(j (;;:,.,.... j, ;, ...;;;; : : .,,,.,,;..: p. ! ,.. Opinion 98 F.

fUNJPmm"., 'I!D .s w,,,,.., Ip,,?, . nce ag,,:l1 rocrer IS I .. , - ' , f I giD wl; i1' 0J1 " r;t :;t ,,!1 ::: f jlr 793. :::JiTI;: J l?t d.r$ jco". 'c";i"' '"79P- . 1 ' Wtt!: 2" ': 51" 1" , ,. . : .; . . ::;; : . , ....__ j, ,,,:; THE KROGER CO. 767 639 Opinion L-"", nul'" I 111,gxs.. :::r TI" U" JI "o" d:=' ,10 it\ r .:::

:Jgj f' J\.

Wii ffS (2, :: I ;7':-!s, 1 : n59' :" 9 l!r I _.

fs;c..

"!f. :'i "I :e_ )11" .. . , ..::_ .:..,.,...,;:,;._ .y.. ,..,::.. .,....... .. .:j::, ?,,,.,.. :...... , ; , ...., ..;!,,: .:..;:::::::::,,, , .:. . :....,,. .. ?/!. .:;....;,., : Opinion 98 F.

W(DIIHIIII r ""f?'f .rs-r' tl1.e Pnce Patrol proves lie ow pnce ealler fim.

r. :: 1" .r, Jl I;J.C";i"Aic GIANT .. /De, CHEERA1!m 1; "u". Jjl. (;J .J : 1f' 00"'0.",. f1 ""C! '79';; ' :Ji99'""f,',".' j;,:'''!8:B'5'I''- 11 t;; lc'CII: !"'I ''''''J. "",, ..I"'1;"'! 3,,- .: -- 1ft Ii. 1 i . l":"-I tJ.:ti 199' 41!) ,.1 li 9S' i;. 1.'1 :,,' 111 5- 1'1 'I.'I" I"'I "" ::I' "II II:\ ::i c ,. iii iS';'1 S7' 19' 1" 69- I;o. I , ".;t:_ n. 'C!2: r49"c""- I - ; i Jj, I j;?iJ -"1" .. ' ;i: ;"".;3-,:;=1..u. ,-: I1,: OIltnla"""' w;;-,. I !nu'-;."""'t I" :: '1 =- 57' I , Tli" KROGER CO. 769 639 Dissenting Statement DISSENTING STA1'EMENT OF COMMISSIONER BAILEY A majority of the Commission has concluded that Kroger violated the Federal Trade Commission Act not because its claims of lower grocery prices through its Price Patrol advertising campaign were false or because the company lacked adequate substantiation for these claims. Rather, according to the majority, Kroger s Price Patrol ads not only asserted that Kroger s prices were generally lower than its competitors but also contained implied claims that the Price Patrol survey was a "methodologically sound" statistically projectible" study which "proved" Kroger prices were lower. The majority concludes that these implied claims were not substantiated. I dissent from the majority s finding of liability in this case because I fundamentally disagree with my colleagues' interpretation of the advertisements at issue: I believe that the ads merely claimed Kroger s prices were lower than its competitors. According to the Commission s opinion and the evidence developed in this case, the general claims of lower prices probably were substantiated by Kroger, although implied claims of a "methodologically sound" statistically projectible" survey were not. Since I do not read the ads to contain such implied claims, I would dismiss the complaint. (2) In addition to my disagreement with the majority s reading of the ads and the standards they have applied in defining adequate substantiation for them, I am moved to dissent from their opinion because this case is also important from a policy perspective. Despite the relatively narrow order entered, I fear that the majority opinion will chil the development of useful comparative price advertising, especially among retail grocers, which is extremely useful to the public, particularly in times of high inflation. I was not a member of the Commission when this complaint was issued and it is clearly too late to revisit that decision. However, it is worth noting, as the majority does, that the Commission s intervention resulted in respondent' s suspension of the Price Patrol or any similar advertising campaign. That result is at least as unfortunate, in my view, as the ultimate finding of liability that the Commission issues today. As the majority opinion implicitly acknowledges, the relationship between the nature of the claims made in an advertisement and the substantiation which will be required for those claims is a dynamic one; generally, the more factual and assertive the Commission Ends the claims, the higher the level of substantiation it wil demand. Recently, the Commission has explored in several cases the presence of implied claims that a factual assertion has been scientifically proven. This type of cleims triggers the highest standards for Dissenting Statement 98 F. adequate substantiation. On the opposite end of the spectrum, a claim which is mere "puffery" may require no substantiation at all. (See, e. , Pfizer, Inc. 81 F. C. 23, 64) Balancing these interactive criteria and distinguishing between the many shades of gray that any particular set of facts may produce is admittedly a difficult and sensi ti ve task.

In the instant case, the majority has derived three basic claims from Kroger s Price Patrol advertisements. First, the majority concludes that the ads only claimed that Kroger had more items with lower prices than its competitors, thereby reversing the Administramost items intive Law Judge s (ALJ) finding that the ads claimed Kroger stores are priced lower. I agree with the majority s holding on this point.

Second, the majority concludes that the challenged advertisements also claimed that shoppers would save overall at Kroger. Once again I agree that the ads at issue made this claim. Finally, the majority holds that the ads conveyed the impression that Kroger had "projectible" and "methodologically sound" statistical surveys to "prove" its lower price claims. I emphatically disagree with that conclusion. (3) The majority s derivation of this "methodologically sound proof' claim occurs in two stages: (1) the Price Patrol surveys were offered as "proof' of the overall savings claim because Hthe commercials invite consumers to assume that the Price Patrol data can be projected reliably to all types of items sold by Kroger; and (2) (iJnherent in the proof claim is the notion that the surveys were conducted in such a way that their results are reliable enough to prove the overall savings representation. This is simply a definition of methodological soundness." (Slip op. at pg. 13) I cannot agree that consumers would ever make the inferential leap suggested by my colleagues concerning the "methodological soundness" of the Price Patrol surveys.

Kroger made no effort to disguise the nature of the Price Patrol; rather, the fact that patrols were composed of homemakers with practical experience in managing a family budget was highlighted in the advertisements. Every effort was made to create a familar, informal atmosphere; the ads sometimes even incorporated comments that a particular Price Patroller s absence from the group was due to a recent childbirth or the family s transfer to another city. Further, the actual Price Patrol survey results were included in the ads in the form of a box score laying out very specifically and clearly the total number of items which had been surveyed, and showing on which Kroger either: (1) "was lower on this many items checked"; (2) THE KROGER CO. 771 639 Dissenting Statement was higher on this many items checked"; or (3) "had the same price s televisionon this many items checked." In most of Kroger commercials, and some of its newspaper advertisements, the company directed consumers' attention to the checklists used by Price Patrollers which were regularly posted in Kroger stores and showed both what items had been surveyed and what prices had been charged. As the majority acknowledges, one Kroger study found that between 8 and 30 per cent of food shoppers who associated the Price Patrol campaign with Kroger had referred to the checklists in the stores. (Slip op. at pg. 5) Although the checklists could not completely cure any deception allegedly inherent in the advertisements, they demonstrate Kroger s effort to delineate the survey s scope accurately.I am convinced that, in dramatic contrast to the majority interpretation, the overall impression actually created by the advertisements was that some homemakers enlisted by Kroger had conducted a limited survey of between 100 and 150 individual grocery items which showed that in most cases Kroger either had the same or lower price than its competitors. The notion that these ansurveys were "methodologically sound" and therefore "prove" anoverall "projection" of lower prices is simply, in my view, artificial construct imposed by the majority on the ads in question. (4) Kroger cites the Commission s decision in Pfizer, Inc., supra support its contention that the "folksiness" of the Price Patrol campaign belies any implied claims of "statistical proof' and methodological soundness." I agree both with respondent' s reading of that ease and its application to the facts at issue here. Finally, I would note that complaint counsel offered no extrinsic proof concerning the meaning of the Price Patrol advertisements and instead relied on the ALJ's and the Commission s expertise to interpret the claims at issue. The ALJ rendered his findings that complaint counsel had met the burden of proving that certain claims were in the ads on a motion for summary judgment. The majority upholds the ALJ's deeision, over respondent's objections, because summary judgment is appropriate when the challenged claims are so clearly conveyed by an advertisement that no genuine issue as their existence can be raised. " (emphasis added, Slip op. at pg. 10) clearly Obviously, I disagree that the ehallenged claims were so conveyed as to remove "any genuine issue" concerning their existence and thus I find the procedural context and resultant evidentiary basis for the majority s opinion on this threshold issue particularly troubling.

. . . . . . . Final Order 98 F.

Having arrived at a different interpretation of the advertisements than the majority, the question remains whether the Price Patrol surveys were adequate substantiation for the claims I believe were contained in the ads. To answer this question, I must turn to the majority opinion:

(WJe agree that the evidence does not dispute Kroger s basic argument that its See RX 925-964. Ifprices were frequently below or equal to those of its competitors. respondent' s advertising had actually been confined to making correspondingly limited claims, the data that it had-the Price Patrol and the Full Book and Burgoyne checks-would have come much closer to providing adequate substantiation. As we have found, however, the Price Patrol claims were considerably more specific, and consequently, a good deal stronger as well. In effect, Kroger made its own substantiating data an integral part of its advertising.. The expert testimony adduced by complaint counsel and respondent ha''i convinced us that the Price Patrol surveys were not methodologically sound and thus did not prove that shoppers would save overall at Kroger. (Expert) endorsements support (5)Kroger s view that it or even lowerhad a reasonable basis to believe its prices were generally as low as, than those of most competitors. But they fall short of confirming what Kroger advertisements actually claimed-that each Price Patrol survey won by Kroger proved it had the lowest prices around. (Slip op. at pgs. 18-19) As I read the majority s opinion, the Price Patrol, Burgoyne and Full Book surveys were close to adequate substantiation for general lower price claims made by Kroger but could not substantiate the implied claim that Kroger had performed a "methodologically sound" survey. Since I do not see the latter claim in the ads, it follows that even under the majority s somewhat restrictive analysis of Kroger s substantiation, the claims I read in the ads were probably substantiated.

My conclusion that no liability should be found here is underscored by my belief that consumers could clearly understand from the face of the ads the limitations of the Price Patrol survey and therefore the limitations of Kroger s substantiation for its general lower price claims. The majority worries that daily grocery shoppers, unlike business executives, may be incapable of understanding the deficiencies of surveys such as those Kroger performed. (Slip Op. at pg. 18) On the contrary, I believe that daily grocery shoppers are at least as capable as business executives of evaluating the clearly disclosed parameters of surveys like the Price Patrol and applying their results to routine purchasing decisions. FINAL ORDER This matter has been heard by the Commission upon the appeal of respondent, and upon briefs and oral argument in support of and in opposition to the appeal. The Commission, for the reasons stated in KHUUEH co. 773 639 Final Order the accompanying Opinion, has granted the appeal in part, and denied the appeal in part. Therefore It is ordered That the initial decision of the administrative law judge be adopted as the Findings of Fact and Conclusions of Law of the Commission, except as is otherwise inconsistent with the attached Opinion.

Other Findings of Fact and Conclusions of Law of the Commission are contained in the accompanying Opinion. It is further ordered That the following Order to Cease and Desist be entered: (2) ORDER A. Respondent means The Kroger Company, a corporation, its successors and assigns, and its officers, representatives, agents and employees, acting directly or indirectly through any corporation, subsidiary or other device in the sale of food, household items and other merchandise in or affecting commerce, as "commerce" is defined in the Federal Trade Commission Act. B. Survey- based food price comparison means an advertised claim that refers to a survey of respondent's and any competitor s food prices and that projects the results obtained from the survey sample to items not included in the survey.

II.

It is ordered That respondent cease and desist from advertising any survey-based food price comparison that refers, directly or indirectly, to a particular city, metropolitan area or competitor (or competitors) by name or other designation unless: A. The survey that forms the basis for the claim is designed and executed in a competent and reliable manner. In addition to such other procedures necessary for competency and reliability, the selection of items for the survey shall be made by respondent or an independent outside firm either:

(i) randomly from the population of products to which the results wil be generalized in the claim; or (ii) in such a way as to make the items representative of the population to which the results wil be generalized in the claim; provided that employees responsible for pricing respondent's mer- Final Order 98 F. chandise do not know which items have been selected for the survey prior to its completion. (3) If selections are made pursuant to subparagraph (ii), the items whose prices are compared should be identical or substantially similar. For the purposes of this paragraph, substantially similar means:

a) branded items which are the lowest priced items available in a product category;

b) meats identical in cut or grade;

cJ produce items of the same type and geographic origin; and B. The claim fairly and impartial1y presents those conclusions that may validly be drawn from the survey and does not gcneralize the results of the survey to a product category that has been systematicaJ1y excluded therefrom; provided, however that no such generalization will be deemed to extend to any product category whose systematic exclusion is disclosed clearly and conspicuously in respondent' s advertisements.

III.

It is further ordered That respondent shall, within sixty (60) days after service of this Order upon it, file with the Commission a written report setting forth in detail the manner and form in which it has complied, or intends to comply, with this Order. IV.

It is further ordered That respondent notify the Commission at least thirty (30) days prior to any proposed change in the corporation such as dissolution, assignment, or sale, resulting in the emergence of a successor corporation, the creation or dissolution of subsidiaries or any other change in the corporation which may affect compliance obligations arising out of this order. It is further ordered That the allegations contained in Paragraphs Six Band D, and Seven Band D, of the Complaint be, and they hereby are, dismissed.

Commissioner Bailey voted in the negative. , KENNECOTT CORP. 775 775 Complaint