B-182105, SEP 21, 1977, 56 COMP GEN 963

B-182105, SEP 21, 1977, 56 COMP GEN 963

CONTRACTS - PRIVITY - SUBCONTRACTORS - LIABILITY FOR CONTRACT OVERPAYMENTS PRIVITY OF CONTRACT DOCTRINE DOES NOT BAR CLAIM BY GOVERNMENT FOR OVERPAYMENTS AGAINST SUBCONTRACTOR WHERE SUBCONTRACTOR BILLED AND ULTIMATELY RECEIVED FROM GOVERNMENT SUBSTANTIALLY ALL OF THE CONTRACT PAYMENTS. DEBT COLLECTIONS - REFERRAL TO JUSTICE - CONTRACT MATTERS - SET-OFF WHERE AMOUNT OF CLAIM ASSERTED BY AGENCY AGAINST SUBCONTRACTOR FOR RECOVERY OF OVERPAYMENTS IS BASED ON STATISTICAL SAMPLING OF 5.6 PERCENT OF ORDERS UNDER CONTRACT RATHER THAN ON AN AUDIT OF EACH CONTRACT ORDER, CLAIM IS NOT SO CERTAIN IN AMOUNT AS TO WARRANT SETOFF BY GENERAL ACCOUNTING OFFICE. HOWEVER, BECAUSE LIABILITY EXISTS, MATTER IS REFERRED TO DEPARTMENT OF JUSTICE FOR APPROPRIATE ACTION.

IN THE MATTER OF THE ARTECH CORPORATION, SEPTEMBER 21, 1977:

THE ARTECH CORPORATION (ARTECH), A SUBCONTRACTOR, HAS APPEALED OUR CLAIMS SETTLEMENT OF JANUARY 18, 1977 (DW-2-2521738), THAT ARTECH IS INDEBTED TO THE UNITED STATES IN THE SUM OF $146,390.00 AS A CONSEQUENCE OF ITS INVOLVEMENT WITH EDUCATIONAL LEARNING SYSTEMS, INC. (ELS), THE PRIME CONTRACTOR, AND THE GENERAL SERVICES ADMINISTRATION (GSA) IN THE PERFORMANCE OF GSA CONTRACT NUMBER GS-01S-4640.

THE CONTRACT, AWARDED ON AUGUST 23, 1970, TO ELS, CALLED FOR THE SUPPLY OF SIX CLASSIFICATIONS OF BOOKS AT PUBLISHERS' LIST PRICES LESS THE DISCOUNTS BID IN EACH OFFER. ELS BID DISCOUNTS WHICH VARIED BY CLASSIFICATION AND QUANTITY AS FOLLOWS:

SPECIAL

CLASSIFICATION NUMBER DISCOUNT

TECHNICAL 36-7 24 TO 30% TEXT 36-8 15 TO 20% TRADE 36-9 37 TO 40% PAPER BOUND 36-10 25 TO 31% MISCELLANEOUS 36-11 10 TO 18% LIBRARY BOUND 36-12 13%

THE CONTRACT TERM BEGAN ON OCTOBER 1, 1970 AND ENDED SEPTEMBER 30, 1971. AUDITS CONDUCTED AFTER COMPLETION OF THE CONTRACT INDICATED THAT MOST OF THE BOOKS SHIPPED HAD BEEN IMPROPERLY CATEGORIZED, WITH THE RESULT THAT THE GOVERNMENT RECEIVED LOWER DISCOUNTS THAN THOSE TO WHICH IT WAS ENTITLED. THE PRINCIPAL REPORTED MISCLASSIFICATION OCCURRED IN THE LIBRARY BOUND CLASSIFICATION WHERE THE GOVERNMENT RECEIVED THE LOWEST DISCOUNT (13 PERCENT). BASED ON A STATISTICAL SAMPLING, GSA HAS DETERMINED THAT THE GOVERNMENT WAS OVERCHARGED IN THE AMOUNT OF $146,390.

THE FILE SHOWS THAT FROM OCTOBER 1, 1970 TO JANUARY 23, 1971, ELS HAD SALES UNDER THE CONTRACT TOTALING $28,539. ON JANUARY 23, 1971, ELS AND ARTECH ENTERED INTO AN AGREEMENT CAPTIONED "SUBCONTRACT," PURSUANT TO WHICH ARTECH WAS TO PERFORM, ON BEHALF OF ELS SUBSTANTIALLY ALL OF ELS' DUTIES UNDER CONTRACT NO. GS-01S-4640. THE DOCUMENT PROVIDED THAT ELS PERSONNEL WOULD REASONABLY ASSIST ARTECH "IN THE PERFORMANCE OF THIS CONTRACT." IT FURTHER PROVIDED THAT ELS, UPON REQUEST OF ARTECH, WOULD COOPERATE WITH ARTECH SO AS TO ENABLE ARTECH TO "QUALIFY AND PERFORM AS A SUBSTITUTE CONTRACTOR OR THE EQUIVALENT, WITH GOVERNMENT APPROVAL, IN THE EVENT OF ELS INSOLVENCY, BANKRUPTCY, DISSOLUTION OR OTHER OCCURRENCE WHICH MIGHT OR DOES RESULT IN A DEFAULT TERMINATION" OF THE ELS CONTRACT. THE AGREEMENT ALSO PROVIDED THAT ELS WOULD ASSIGN MONIES DUE UNDER THE CONTRACT TO A FINANCING INSTITUTION AS MIGHT BE DESIGNED BY ARTECH.

ELS THEN REQUESTED THAT THE GSA CONTRACTING OFFICER MODIFY THE CONTRACT BY CHANGING THE NAME AND ADDRESS OF THE CONTRACTOR TO READ "EDUCATIONAL LEARNING SYSTEMS/ARTECH DIVISION, ARTECH CORP." AT ARTECH'S ADDRESS. THE CONTRACTING OFFICER ADVISED ELS THAT "THIS CONTRACT CANNOT BE ASSIGNED AS PROPOSED," BUT DID ISSUE A CONTRACT MODIFICATION CHANGING THE MAILING ADDRESS OF THE CONTRACTOR TO THAT OF ARTECH. ARTECH COMPLETED PERFORMANCE OF THE CONTRACT, WITH CONTRACT SALES OF $808,967. PURSUANT TO THE TERMS OF THE SUBCONTRACT ARTECH RECEIVED A POWER OF ATTORNEY WHICH ENABLED IT TO CASH GOVERNMENT CHECKS REPRESENTING CONTRACT PAYMENTS MADE OUT TO ELS. ARTECH CONTINUED TO RECEIVE AND CASH THE GOVERNMENT CHECKS UNTIL JUNE 1971 WHEN PAYMENTS WERE DIVERTED TO ELS' ASSIGNEE FOR THE BENEFIT OF CREDITORS. HOWEVER IN AUGUST 1971 PURSUANT TO A COURT ORDER ARTECH ONCE AGAIN BEGAN TO RECEIVE THE PROCEEDS FLOWING FROM ITS PERFORMANCE OF THE SUBCONTRACT.

IN THE INTERIM, IN JUNE 1971, ELS EXECUTED AN ASSIGNMENT FOR THE BENEFIT OF CREDITORS AND CEASED OPERATIONS AS A VIABLE CONCERN. AT ABOUT THAT TIME, ARTECH AND THE ELS ASSIGNEE, BELIEVING THAT THE GSA CONTRACT WAS IN DANGER OF TERMINATION FOR DEFAULT AND IN ORDER TO RESOLVE DISPUTES CONCERNING THE SUBCONTRACT, WHICH HAD RISEN BETWEEN ARTECH AND ELS, ENTERED INTO A COMPROMISE AGREEMENT WHICH WAS RATIFIED BY THE CIRCUIT COURT OF MONTGOMERY COUNTY, MARYLAND, ON AUGUST 4, 1971. PURSUANT TO THE AGREEMENT AND THE COURT'S ORDER, ARTECH WAIVED ALL CLAIMS IT MIGHT HAVE AGAINST ELS ARISING OUT OF THE JANUARY 22, 1971 SUBCONTRACT, AGREED TO FAITHFULLY PERFORM UNDER THE TERMS OF THE SUBCONTRACT, AGREED "TO HONOR ITS COMMITMENTS FOR PAYMENT OF PRIME CONTRACT PAYABLES ASSUMED UNDER THE SUBCONTRACT *** AS THE SAME ARE PROPERLY PRESENTED TO IT," AND AGREED "TO IDEMNIFY AND HOLD HARMLESS" THE ASSIGNEE AND ESTATE OF ELS "FROM ANY LIABILITY ARISING OUT OF ACTS OR FAILURES TO ACT ON THE PART OF ARTECH CORP. IN CONNECTION WITH ITS PERFORMANCE UNDER THE SAID SUBCONTRACT OF JANUARY 22, 1971."

GSA'S APRIL 1, 1974 AUDIT OF THE CONTRACT DISCLOSED THAT IN ONLY 35 ORDERS, OUT OF THE STATISTICAL SAMPLE OF 120 ORDERS EXAMINED (A TOTAL OF 2,136 ORDERS WERE PLACED), DID THE FEDERAL ORDERING AGENCY SPECIFY THE CLASSIFICATION OF THE BOOKS SOUGHT. THUS IT APPEARS THAT IN NUMEROUS INSTANCES ARTECH SELECTED THE DISCOUNT RATE WHICH WOULD BE APPLIED TO A PARTICULAR ORDER.

GSA HAS TAKEN THE POSITION THAT ARTECH IN MANY INSTANCES SELECTED THE WRONG DISCOUNT AND THAT THE GOVERNMENT IS ENTITLED TO A REFUND FROM ARTECH FOR THE RESULTING OVERCHARGES. IT BELIEVES THAT ARTECH BECAME, IN EFFECT, THE PRIME CONTRACTOR AND THAT THE GOVERNMENT IS ENTITLED TO RECOVER FROM ARTECH BASED ON THE THEORY OF EQUITABLE ESTOPPEL OR ON A THEORY OF AGENCY. OUR CLAIMS DIVISION AGREED WITH GSA, STATING THAT THE PARTICULAR RELATIONSHIP BETWEEN ARTECH/ELS AND THE UNITED STATES GIVES RISE TO A DIRECT LIABILITY OF ARTECH BASED ON A THIRD PARTY BENEFICIARY THEORY ALONG WITH AN AGENCY THEORY. CITING KERN-LIMERICK V. SCURLOCK, 347 U.S. 110 (1954); DELTEC CORP. V. UNITED STATES, 326 F.2D 1004, 164 CT.CL. 432 (1964); AND 21 COMP.GEN. 682 (1942) OUR CLAIMS DIVISION CONCLUDED THAT THE CIRCUMSTANCES OF THIS CASE "CLEARLY GIVE RISE TO THESE EXTRAORDINARY THEORIES OF LIABILITY."

ARTECH DISAGREES WITH THE LEGAL THEORY THAT ARTECH, A SUBCONTRACTOR, IS LIABLE TO THE GOVERNMENT FOR ANY OVERCHARGES UNDER THE PRIME CONTRACT. ARGUES THAT THROUGHOUT PERFORMANCE OF THE CONTRACT GSA INSISTED THAT ARTECH WAS ONLY A SUBCONTRACTOR AND THAT THE GOVERNMENT'S DEALINGS MUST BE WITH ELS. THEREFORE, ARTECH ARGUES, "ON THE FACTS IT MUST BE DETERMINED THAT THE GOVERNMENT NEGATED ANY THIRD-PARTY BENEFICIARY OR AGENCY RELATIONSHIP WITH ARTECH."

MOREOVER, ARTECH ARGUES THAT EVEN IF THE GSA/GAO CLAIMS DIVISION THEORY OF LIABILITY IS CORRECT, ARTECH COULD NOT BE HELD LIABLE TO THE GOVERNMENT FOR ANY OVERCHARGES PRIOR TO AUGUST 1971, "WHEN THE MONTGOMERY COUNTY CIRCUIT COURT FIRST ORDERED THAT SALES ON THIS CONTRACT WERE NOT TO BE RUN THROUGH THE RECEIPTS OF THE ASSIGNEE FOR THE BENEFIT OF CREDITORS OF ELS."

AS ARTECH POINTS OUT, RECOVERY UNDER A CONTRACT IS GENERALLY LIMITED TO PARTIES IN PRIVITY WITH EACH OTHER AND NORMALLY THERE IS NO PRIVITY OF CONTRACT BETWEEN THE GOVERNMENT AND SUBCONTRACTORS. MERRITT V. UNITED STATES, 267 U.S. 338 (1925). THE ABSENCE OF PRIVITY, HOWEVER, WILL NOT DEFEAT RECOVERY IF THE CIRCUMSTANCES INDICATE THAT THE RELATIONSHIP BETWEEN THE PARTIES WAS SOMETHING OTHER THAN THE NORMAL GOVERNMENT- SUBCONTRACTOR RELATIONSHIP. SEE KERN-LIMERICK V. SCURLOCK, SUPRA; DELTEC CORP. V. UNITED STATES, SUPRA; AND 21 COMP.GEN. 682, SUPRA (WHERE THE PRIME CONTRACTOR ACTED AS AGENT OF THE GOVERNMENT); UNITED STATES V. HUFF, 165 F.2D 720 (5TH CIR. 1948) AND MANEELY V. UNITED STATES, 68 CT.CL. 623 (1929) (WHERE THE SUBCONTRACTOR WAS CONSIDERED TO BE A THIRD PARTY BENEFICIARY OF THE GOVERNMENT CONTRACT); UNITED STATES V. GEORGIA MARBLE CO., 106 F.2D 955 (5TH CIR. 1939) AND B-175550, JUNE 14, 1973 (WHERE THE GOVERNMENT'S ACTUAL OR IMPLIED PROMISE TO PAY RESULTS IN SUBCONTRACTOR PERFORMANCE). THAT AN AGENCY RELATIONSHIP MAY EXIST BETWEEN A PRIME CONTRACTOR AND ANOTHER PARTY, EVEN THOUGH THAT PARTY IS REFERRED TO AS A SUBCONTRACTOR, HAS BEEN RECOGNIZED BY BOTH THE COURTS AND THE BOARDS OF CONTRACT APPEALS. HUNT V. UNITED STATES, 257 U.S. 125 (1921); GLENS FALLS INSURANCE CO. V. NEWTON LUMBER & MFG. CO., 388 F.2D 66 (10TH CIR. 1967); APPEAL OF CENTRAL MACHINE & TOOL CO., ASBCA NO. 837, JUNE 13, 1952. FINALLY, WHERE ALL THE ESSENTIAL ELEMENTS TO ESTABLISH EQUITABLE ESTOPPEL ARE PRESENT, SEE UNITED STATES V. GEORGIA-PACIFIC CO., 421 F.2D 92, 96 (9TH CIR. 1970), A SUBCONTRACTOR MAY BE ESTOPPED FROM DENYING THAT IT WAS THE PRIME CONTRACTOR.

ARTECH MAINTAINS THAT THE CASES SUPPORT ITS POSITION OF NO LEGAL LIABILITY TO THE GOVERNMENT AND IT SPECIFICALLY CITES HUNT V. UNITED STATES, SUPRA, AND GRAY & CO. V. UNITED STATES, 79 CT.CL. 117 (1934) IN THIS REGARD. IN HUNT THE SUPREME COURT HELD THAT A PRIME CONTRACTOR COULD RECOVER FROM THE GOVERNMENT FOR EXTRA SERVICES PERFORMED ALTHOUGH THE SERVICES HAD BEEN PERFORMED BY A SUBCONTRACTOR. WE DO NOT READ THIS CASE AS SUPPORT FOR ARTECH'S POSITION; IN FACT THE COURT RECOGNIZED THAT THE RELATIONSHIP BETWEEN THE PRIME AND SUBCONTRACTOR WAS TREATED BY THEM AS ONE OF AGENCY. GRAY INVOLVED A CASE WHERE THE GOVERNMENT TERMINATED A CONTRACT FOR CONVENIENCE AND IN CONNECTION WITH THE TERMINATION SETTLEMENT PAID THE PRIME CONTRACTOR AND THEN MISTAKENLY PAID THE SUBCONTRACTOR FOR THE SAME MATERIAL WHICH HAD BEEN FURNISHED BY THE PRIME TO THE SUBCONTRACTOR TO PERFORM THE CONTRACT. THE GOVERNMENT ATTEMPTED TO RECOVER THE DOUBLE PAYMENT FROM THE PRIME CONTRACTOR, THE SUBCONTRACTOR HAVING GONE OUT OF EXISTENCE. THE COURT DENIED RECOVERY, STATING THAT WHILE THE GOVERNMENT PAID TWICE FOR THE SAME MATERIAL, "THIS DOES NOT JUSTIFY THE RECOVERY OF THE AMOUNT FROM THE WRONG PARTY, OR THE INNOCENT PARTY, AND THE ONLY ONE FROM WHOM COLLECTION CAN BE MADE." IN OUR OPINION, THIS CASE SUPPORTS THE GSA VIEW THAT WHERE AN ERRONEOUS PAYMENT IS MADE BY THE GOVERNMENT TO A SUBCONTRACTOR, RECOVERY SHOULD BE SOUGHT FROM THE SUBCONTRACTOR AND NOT FROM THE PRIME CONTRACTOR.

HERE, ARTECH, ALTHOUGH DENOMINATED A SUBCONTRACTOR, WAS ESSENTIALLY AUTHORIZED BY ELS TO TAKE OVER THE GSA CONTRACT, TO PERFORM IT IN ACCORDANCE WITH THE CONTRACTUAL PROVISIONS AND APPLICABLE LAWS AND REGULATIONS AND, BY VIRTUE OF THE POWER OF ATTORNEY EXECUTED SUBSEQUENT TO THE SUBCONTRACT AGREEMENT, TO ACCEPT CONTRACT PAYMENTS MADE IN THE NAME OF ELS. MOREOVER, THE RECORD INDICATES THAT ARTECH RETAINED ALL MONIES IT RECEIVED PURSUANT TO THE CONTRACT, INCLUDING OVERPAYMENTS. WE BELIEVE THAT THE GOVERNMENT HAS A VALID LEGAL CLAIM AGAINST ARTECH FOR ANY AND ALL OVERPAYMENTS WHICH WERE RECEIVED BY ARTECH. WE THINK IT IS CLEAR FROM THE CASES THAT THE "NO PRIVITY" RULE WILL NOT STAND IN THE WAY OF RECOVERY, BY EITHER THE GOVERNMENT OR BY A SUBCONTRACTOR AGAINST THE GOVERNMENT, WHERE THE CIRCUMSTANCES JUSTIFY RECOVERY.

IN THIS CONNECTION, WE DO NOT AGREE WITH ARTECH THAT IT SHOULD NOT BE HELD ACCOUNTABLE FOR OVERPAYMENTS RECEIVED PRIOR TO AUGUST 1971. WHILE ARTECH REPORTS THAT CONTRACT SALES RECEIPTS RECEIVED AFTER THE JUNE 1971 ASSIGNMENT FOR THE BENEFIT OF CREDITORS AND PRIOR TO THE AUGUST 1971 COURT ORDER WERE TURNED OVER TO THE ASSIGNEE, GSA STATES THAT THE OVERCHARGE WAS COMPUTED ON THE BASIS OF SALES FOR WHICH PAYMENT WAS ULTIMATELY RECEIVED BY ARTECH. IT REPORTS THAT AFTER ELS HAD MADE AN ASSIGNMENT FOR THE BENEFIT OF CREDITORS, CHECKS IN THE AMOUNT OF $10,587.36 WERE RECEIVED BY ELS AND TURNED OVER TO THE ASSIGNEE, WHO KEPT 10 PERCENT, OR $1,058.74, AND TRANSMITTED THE BALANCE TO ARTECH. THUS, GSA STATES THAT IF AN ADJUSTMENT FOR ANY RECEIPTS NOT GIVEN TO ARTECH BY THE ASSIGNEE IS REQUIRED, THEN 18.0959 PERCENT (THE OVERCHARGE RATE DETERMINED BY GSA) OF $1,058.74, OR $191.59 SHOULD BE DEDUCTED, LEAVING $146,198 ($146,390 LESS $191.59) AS THE OVERCHARGE. WE SEE NO REASON TO DISAGREE WITH THIS ANALYSIS.

ARTECH ALSO ATTACKS THE GSA FINDING AS TO THE AMOUNT OF THE OVERCHARGES. ARTECH POINTS OUT THAT THE QUESTION OF AMOUNT OWED WAS DETERMINED IN THIS CASE BY THE CLASSIFICATION, OF THE BOOKS ORDERED, AND THAT CLASSIFICATIONS WERE "NOT EASY TO DETERMINE BECAUSE OF THE OVERLAPPING DESCRIPTIONS IN THE SPECIFICATIONS." IT POINTS OUT, FOR EXAMPLE, THAT THE RECENT BEST SELLER, "ROOTS," COULD CONCEIVABLY BE CLASSIFIED AS A TECHNICAL BOOK, DUE TO ITS TECHNICAL OR SCIENTIFIC NATURE, A TEXT BOOK, BECAUSE IT IS EDUCATIONAL, A TRADE BOOK, BECAUSE IT DOES HAVE GENERAL INTEREST AND BIOGRAPHICAL MATTER, AND THE DISCOUNT WOULD VARY DEPENDING UPON THE CLASSIFICATION. IT ARGUES THAT THE "CONTRACT DOES NOT HAVE A CLAUSE AKIN TO A WARRANTY CLAUSE WHICH BINDS THE CONTRACTOR TO A RE-EVALUATION THREE YEARS AFTER DELIVERY AND ACCEPTANCE OF BOOKS BY SO-CALLED 'LIBRARY EXPERTS' - WHICH IS WHAT THE 1974 (GSA) AUDIT IS BASED ON."

MOREOVER, ARTECH OBJECTS TO THE "STATISTICAL SAMPLE" APPROACH USED BY GSA TO DETERMINE THE EXTENT OF OVERCHARGES. IT NOTES THAT OUT OF A TOTAL OF 2,136 ORDERS, 120 WERE EXAMINED AND ALL RESULTS WERE EXTRAPOLATED FROM THIS SAMPLE. YET, ARTECH STATES, THE ORDERS WERE "NEITHER TANGIBLE NOR IDENTICAL UNITS," SINCE EACH ORDER VARIED SUBSTANTIALLY IN TERMS OF VOLUMES AND TITLES. ARTECH ARGUES THAT WHILE THE GSA SAMPLE REPRESENTS ABOUT 5.6 PERCENT OF THE ORDERS PLACED, "THE NUMBER OF VOLUMES ON THOSE ORDERS COULD THEORETICALLY HAVE BEEN LESS THAN 1 PERCENT OF ALL VOLUMES ORDERED, AND CERTAINLY NOT EXCEEDING 2 PERCENT." IN FACT ARTECH STATES THAT IT RE-EXAMINED THREE OF THE 120 ORDERS COVERED IN THE GSA SAMPLE, AND IT FOUND ONLY MINOR OVERCHARGES, MUCH LESS THAN THE AMOUNTS DETERMINED IN THE GSA AUDIT. THE RECORD SHOWS THAT GSA PERFORMED TWO AUDITS OF THE CONTRACT. THE FIRST AUDIT REPORT, DATED JANUARY 22, 1973, FOCUSED ON THE FAILURE OF THE FEDERAL AGENCIES TO CLEARLY STATE THEIR REQUIREMENTS WHEN ORDERING UNDER THE CONTRACT. THE ORDERING ACTIVITIES OFTEN DID NOT DESIGNATE THE CLASSIFICATION OF THE BOOKS SOUGHT WHEN PLACING THEIR ORDERS. THIS LEFT THE CONTRACTOR AT LIBERTY TO DESIGNATE THE CLASSIFICATIONS WHICH CLASSIFICATIONS IN TURN DETERMINED THE DISCOUNTS APPLICABLE TO THE RESPECTIVE ORDERS. IN ORDER TO ASCERTAIN THE IMPACT ON CONTRACT PERFORMANCE, THE GSA AUDITORS ATTEMPTED TO RELATE THE TYPES OF BOOKS THE CONTRACTOR WAS PURCHASING FROM THE VARIOUS PUBLISHERS WITH THE TYPES OF BOOKS BEING DELIVERED TO THE AGENCIES. THEY DISCOVERED THAT ORDERS WERE PLACED WITH APPROXIMATELY 1,200 DIFFERENT PUBLISHERS WITHOUT REFERENCE TO TYPE OF CLOTHBOUND BOOK BEING ORDERED AND THAT OF A SAMPLE OF PUBLISHERS' INVOICES FROM 43 OF THE LARGER ORDERS ONLY SIX INDICATED THE TYPE OF BOOK BEING SUPPLIED. THE AUDITORS WERE THEREFORE UNCERTAIN AS TO THE EXACT NATURE OF THE BOOKS WHICH THE PUBLISHERS HAD FURNISHED THE CONTRACTOR.

SIMILAR ATTEMPTS WERE MADE TO RELATE PUBLISHERS' INVOICES TO THE FEDERAL AGENCY ORDERS WHICH WERE PLACED WITH THE CONTRACTOR BUT THIS PROVED FRUITLESS BECAUSE THE CONTRACTOR'S ACCOUNTING SYSTEM DID NOT CROSS- REFERENCE THE PUBLISHERS' ORDER FILES TO THE AGENCY ORDER FILES. FINALLY THE AUDITORS COMPUTED AN ESTIMATED AMOUNT OF OVERCHARGE BY COMPARING THE PRIOR SALES HISTORY OF DIFFERENT CLASSIFICATIONS OF BOOKS, AS REPORTED BY PRIOR CONTRACTORS, TO THE SALES HISTORY WHICH THE CONTRACTOR CLAIMED TO HAVE EXPERIENCED. ON THIS BASIS THE AUDITORS FOUND AN INDICATED OVERCHARGE OF $87,948.

THE SECOND GSA AUDIT REPORT OF APRIL 1, 1974, USED THE FOLLOWING METHODOLOGY IN OBTAINING AN ESTIMATE OF THE AMOUNT THE GOVERNMENT HAD BEEN OVERCHARGED:

WE OBTAINED TECHNICAL ASSISTANCE FROM LIBRARIANS IN THE NATIONAL ARCHIVES LIBRARY, NATIONAL ARCHIVES AND RECORDS SERVICE. WE LEARNED THAT GENERALLY PUBLISHERS DO NOT USE LIBRARY BINDING ON THE MAJORITY OF THEIR TECHNICAL, TEXT, OR TRADE BOOKS. SOME PUBLISHERS DO NOT OFFER ANY LIBRARY BOUND BOOKS. ***

WE DETERMINED THAT THERE WERE 2,136 ORDERED BOOKS RECEIVED UNDER THE CONTRACT. WE OBTAINED A STATISTICAL SAMPLE OF 120 ORDER NUMBERS AND EXTRACTED THOSE FILES FOR REVIEW. THE NARS LIBRARIANS EXAMINED THE INVOICES AND THE AGENCIES' ORDERS. THEY DETERMINED WHICH BOOKS WERE INCLUDED ON THE INVOICES AND THEN VERIFIED THE CLASSIFICATIONS OF THOSE BOOKS. WE RECOMPUTED THE INVOICES TO PROVIDE FOR THE CORRECT DISCOUNTS BASED ON THE LIBRARIANS' CLASSIFICATION OF THE BOOKS. WE FOUND THAT THE INVOICES EXAMINED TOTALLED $56,836.63 AND WERE OVERSTATED BY $10,285.09 DUE TO THE CONTRACTOR'S FAILURE TO ALLOW THE CORRECT DISCOUNTS.

USING THIS METHODOLOGY, THE AUDITORS CONCLUDED, "WITH A 90 PERCENT CONFIDENCE LEVEL, THAT THE TOTAL OVERCHARGE AMOUNTED TO $146,390.00 PLUS OR MINUS $22,121.20."

BASED ON THE FOREGOING WE CANNOT AGREE WITH ARTECH THAT THE GSA AUDIT FINDINGS WERE INVALID. GSA REPORTS THAT FOR THE MAJORITY OF BOOKS ORDERED ONLY A 13 PERCENT DISCOUNT WAS ALLOWED. THIS IS THE DISCOUNT RATE APPLICABLE TO LIBRARY BOUND BOOKS. THE JANUARY 1973 GSA AUDIT REPORT ESTIMATED THAT ABOUT 71 PERCENT OF THE BOOKS ORDERED DURING THE ENTIRE CONTRACT PERIOD WERE CLASSIFIED AS BEING LIBRARY BOUND. YET, ACCORDING TO GSA, MANY OF THESE BOOKS WERE NOT EVEN OFFERED IN LIBRARY BOUND EDITIONS BY THE PUBLISHERS. MOREOVER, GSA STATES THAT BASED ON PRIOR CONTRACT ORDERS ONLY ABOUT 6.5 PERCENT OF THE TOTAL BOOKS ORDERED WERE LIBRARY BOUND.

WE NOTE THAT LIBRARY BOUND REFERS TO THE PHYSICAL NATURE OF THE BOOK ITSELF, UNLIKE MOST OF THE OTHER CLASSIFICATIONS. THUS A BOOK CAN HAVE A TRADE SUBJECT MATTER ENTITLING THE GOVERNMENT TO A 37 TO A 40 PERCENT DISCOUNT AND AT THE SAME TIME BE LIBRARY BOUND, WHICH ONLY ENTITLES THE GOVERNMENT TO A 13 PERCENT DISCOUNT. WHILE THE CONTRACT ITSELF HAD NO PROVISION TO COVER SUCH OVERLAPPING CLASSIFICATION, WE THINK IT IS REASONABLE TO CONCLUDE THAT THE CONTRACTOR MAY CLASSIFY A BOOK AS LIBRARY BOUND IN THE SITUATION DESCRIBED ABOVE.

IN THIS CONNECTION, LIBRARY BINDING, AS WE UNDERSTAND IT, MEANS A BINDING STRONGER THAN THAT WHICH WOULD ORDINARILY BE FURNISHED. WEBSTER'S THIRD NEW INTERNATIONAL DICTIONARY, UNABRIDGED, 1966 ED., DEFINES "LIBRARY BINDING" AS "AN ESP. STRONG DURABLE CLOTH BOOK BINDING SUITABLE FOR USE BY A CIRCULATING LIBRARY ***." IN AN OTHERWISE UNRELATED PORTION OF THE SOLICITATION (WHICH ELS DID NOT BID), REFERENCE IS MADE TO "TRADE BOOKS TO BE LIBRARY BOUND." THAT SECTION OF THE SOLICITATION SOUGHT BIDS FOR REBUILDING AND UPGRADING BOOKS (ORIGINALLY ISSUED WITH A TRADE OR EDITION BINDING) TO THE STATUS OF LIBRARY BOUND BOOKS. THE REFERENCED PORTION OF THE SOLICITATION FURTHER INDICATES THAT THE RESTORATION WORK WAS TO MEET THE STANDARDS SET BY THE LIBRARY BINDING INSTITUTE OF BOSTON, MASSACHUSETTS (LBI). LBI HAS ADVISED THIS OFFICE THAT ITS SPECIFICATION FOR CLASS A BINDING, OR LIBRARY BINDING, ORIGINATED WITH LIBRARIANS' DESIRES FOR BINDINGS WHICH WERE MORE DURABLE THAN THE PUBLISHER'S TRADE BINDINGS. THE LIBRARIES PURCHASED TRADE BOUND EDITIONS WHICH WERE THEN CIRCULATED UP TO TEN TIMES BEFORE BEING SENT TO BE REBOUND IN A STRONGER OR "LIBRARY BINDING." CERTAIN PUBLISHERS THEN BEGAN THE PRACTICE OF "PREBINDING" THEIR BOOKS, ESPECIALLY CHILDREN'S BOOKS. ACCORDING TO LBI THE TERM PREBINDING OR PREBOUND IS APPLIED TO NEW BOOKS BOUND ACCORDING TO THE CLASS A STANDARD. IT THUS APPEARS THAT WHEN THE CONTRACT SPECIFIED A CERTAIN DISCOUNT FOR LIBRARY BOUND BOOKS IT INDICATED THAT SUCH DISCOUNT WOULD APPLY TO BOOKS WHICH WERE IN SOME WAY HELD OUT TO THE PUBLIC BY THE PUBLISHERS AS BEING A REINFORCED VERSION OF THE USUAL TRADE BOUND BOOK.

ARTECH ARGUES THAT THE GSA STATISTICAL SAMPLE WAS NOT REPRESENTATIVE. THE STRENGTH OF ITS OWN EXAMINATION OF "THREE RANDOMLY-SELECTED ORDERS" ARTECH FINDS THAT, AT MOST, "THE GOVERNMENT HAS A MAXIMUM OVERCHARGE OF ABOUT $16,000 AND NOT $143,000." THE $16,000 OVERCHARGE, ARTECH STATES, IS BASED ON THE DIFFERENCE BETWEEN THE 13 PERCENT DISCOUNT APPLICABLE TO LIBRARY BOUND BOOKS AND THE DISCOUNT APPLICABLE TO EACH BOOK ORDERED UNDER ONE SAMPLE ORDER EXAMINED BY ARTECH (CLARK AFB ORDER NO. 1680). SINCE ONLY A "MAJORITY" OF THE BOOKS ORDERED ALLEGEDLY WERE MISCLASSIFIED, ARTECH STATES THAT THE TOTAL OVERCHARGE SHOULD BE EVEN LESS THAN $16,000.

WE NOTE THAT IN SAMPLE ORDER NO. 1680, ARTECH CLASSIFIED "CUBA SOCIALISM & DEVELOPMENT," "EXOTIC FANTASIES," "SHORT HISTORY OF CHINESE ART," AND "AGRICULTURAL FORESTING IN OCEAN TECHNOLOGY," AS ALL BEING TEXT BOOKS, SUBJECT TO A 15 PERCENT DISCOUNT RATE SINCE ONLY ONE VOLUME OF EACH WAS ORDERED. THE CONTRACT DEFINED TEXT BOOKS AS EDUCATIONAL, SCHOOL OR REFERENCE BOOKS, AND A DISCOUNT OF 15 TO 20 PERCENT WAS APPLICABLE TO SUCH BOOKS, DEPENDING UPON THE VOLUME OF THE ORDER. TRADE BOOKS WERE DEFINED AS BOOKS OF GENERAL INTEREST, INCLUDING WORKS OF FICTION, BIOGRAPHIES AND GENERAL TITLES WIDELY READ BY THE GENERAL PUBLIC. A DISCOUNT OF 37 TO 40 PERCENT WAS APPLICABLE FOR THESE BOOKS. TECHNICAL BOOKS WERE THOSE DESIGNATED BY PUBLISHERS AS HAND BOOKS AND OTHER PRACTICAL WORKS OF A TECHNICAL, SCIENTIFIC OR BUSINESS NATURE, AND WERE SUBJECT TO DISCOUNTS OF 24 TO 30 PERCENT. WHILE ARTECH HAS CATEGORIZED THE AFOREMENTIONED FOUR TITLES AS TEXT BOOKS (EDUCATIONAL) WE THINK THESE BOOKS COULD MORE REASONABLY BE CLASSIFIED BY GSA AS BEING OTHER THAN TEXT BOOKS, SUCH AS TRADE OR TECHNICAL BOOKS, AND THUS SUBJECT TO THE LARGER DISCOUNTS. ALSO WE NOTE THAT ORDER 1680 ONLY ENTITLED THE GOVERNMENT TO THE MINIMUM DISCOUNT FOR EACH CLASSIFICATION SINCE ONLY ONE OR TWO VOLUMES OF EACH TITLE WERE ORDERED. IN THIS RESPECT WE CANNOT SAY THAT THE ORDER WAS TYPICAL OF THE OTHER ORDERS. IN SHORT, THE EVIDENCE FURNISHED BY ARTECH DOES NOT SHOW THAT THE GSA STATISTICAL SAMPLING OF 120 ORDERS WAS NOT REPRESENTATIVE OF ALL 2,136 ORDERS.

FINALLY, ARTECH ARGUES THAT GSA'S METHOD OF ESTIMATING THE AMOUNT OF THE ASSERTED OVERPAYMENT IS LEGALLY IMPROPER. IT MAINTAINS THAT THE OVERCHARGE DETERMINATION PROPERLY MUST BE BASED ON AN AUDIT OF EVERY SINGLE ORDER UNDER THE CONTRACT AND NOT ON A PROJECTION OF A STATISTICAL SAMPLE OF 5.6 PERCENT OF THE TOTAL ORDERS. WE FIND THAT PRECEDENT DOES EXIST FOR THE USE OF SAMPLE DATA AS EVIDENCE BEFORE ADMINISTRATIVE TRIBUNALS AS WELL AS IN COURT PROCEEDINGS. APPARENTLY COURTS ALTERNATIVE METHOD OF PROOF AND THERE IS PRECEDENT FOR THE USE OF SUCH MAY ACCEPT VALID SAMPLE EVIDENCE AS TO OBJECTIVE FACTS IF THERE IS NO EVIDENCE IN THE PARTICULAR FIELD IN QUESTION. SEE SPROULS, "THE ADMISSIBILITY OF SAMPLE DATA INTO A COURT OF LAW. A CASE HISTORY," 4 U.C.L.A. L. REV. 222, 223 (1957). AS THAT ARTICLE AND A COMPANION ARTICLE INDICATE, "THE COURTS ARE NOT UNWILLING TO MAKE SOME USE OF SAMPLING TECHNIQUES" BUT ARE HESITANT "TO EXTEND THE USE OF SUCH TECHNIQUE BEYOND THE VERY SIMPLEST SAMPLES OF TANGIBLE OBJECTS." SEE MCCOID, "THE ADMISSIBILITY OF SAMPLE DATA INTO A COURT OF LAW: SOME FURTHER THOUGHTS," 4 U.C.L.A. L. REV. 233, 247 (1957). PROFESSOR MCCOID BELIEVES, HOWEVER, THAT STATISTICAL ANALYSIS WILL BE OF MOST USE IN THE COMPLEX TYPE CASES, SUCH AS THOSE INVOLVING ANTITRUST PROBLEMS OR CASES INVOLVING DETERMINATION OF AVERAGE PRICES OVER LONG PERIODS OF TIME, AND HE HOPES THAT THE COURTS CAN BE PERSUADED THAT "RANDOM SAMPLING TECHNIQUES ARE RELATIVELY TRUSTWORTHY, PROVIDED AN APPROXIMATELY LARGE SAMPLE IS SELECTED." BE THAT AS IT MAY, WE CAN FIND NO CLEAR PRECEDENT ANALOGOUS TO THE PRESENT SITUATION WHERE SAMPLE DATA WAS USED FOR THE PURPOSE OF PROJECTING THE AMOUNT OF OVERPAYMENTS UNDER A CONTRACT; NOR HAS GSA CITED ANY PRECEDENT IN ITS REPORT. INDEED GSA ACKNOWLEDGES THE CONJECTURAL NATURE OF ITS CALCULATION.

THEREFORE, WE CANNOT SAY THAT THE GOVERNMENT'S CLAIM IS SO CERTAIN IN AMOUNT AS TO WARRANT SETOFF. 4 C.F.R. SEC. 102.3 (1977). HOWEVER, FOR THE REASONS INDICATED, WE THINK LIABILITY EXISTS AND WE ARE THEREFORE REFERRING THIS MATTER TO THE DEPARTMENT OF JUSTICE FOR APPROPRIATE ACTION.