DOCUMENT RESUMEe

05030 - [B0:1452181 ](Res1fricted)

[ Review cf Proposal ~r Establishing a'Ce'ntralized GovernmentTravel Agency]. LCD-7__29; B-103315. Pebruary 3, 1978.'&'11 pp.

Report to Sen. Edward- orinsky; by Pred J. Shafcr, Director,Loqistics and Commun'iJions Div.

Contact: Logistics and__raaunications Div.Budget Function: Comf4iSc' and Transportation: Air Transportation

(405).Organization Concerne4,'General Services A&ministration.Congressional Relevance: Sen. Edward Zorinsky.Authority: Privacy &ct:. 1974.

A Gorernmentfl:e automated irfine reservation andticketing system was proposed to reduce costs ty using .tp.erpeople. The basis fokit"ings in the Fropoal -is the reductionof personnel curreatfji''j ocessing travel reservaticns .'(diectlabor) and the elimi'nation of indirect labor, such a.---secretaries, involv..the travel process. The savi'ng.iD.indirect labor would.,A.tbe realized by .eliminatin'g an '-personnel but would ;QPF sent incremental savings of tiaLwhichcould be devoted to m.oe.productive ftictians. In addition,overpayments to airlsneTbelieved to occur in current procedurescould be elieinated.'"i~iFtent and proposed. cost estiates-.cannotbe validated because thr'"basic data used in the proposal-wereunreliable. CentralizWst ystea equipaent.costs were estiaated tobe about S960,000 beajsn the proposl's estisate.aof tEavelvolume. This estimfati" presliminl.ari'Q wia apparentl asedon a misunderstanding -o the syst:a' s conce'pt'. 'ba'doniag indveendently develop4d.agen'c7 'reseruation and ticketing-syTstemsin tavor of a centralized -sstems eq'uiesi planning-. ' -coordination,- and coneinss,. The proposed. system aplpieatrs to betechnically feasible-.beause of its-similarity to site'ei sed -"aby major airlines'. Rouwer, alternative mithods 'have not beenconsidered. (RBS) -. X ;:-. .. ::

UNITED STATES GENERAL ACCOUNTING OFFICE

D iot tC ba e. t4 ,of,,,df Cjt. 8

B-103315 FEB 0 3 1978

The Honorable Edward ZorinskyUnited States Senate

Dea: Senator Zorinsky:

In your April 7, 1977, letter, you asked us to evaluatea proposal to establish a centralized Government travelagency. Specifically, you asked that we consider costsand the personnel savings which could be realized by adopt-ing the propcsal, together with any other collateral bene-fits and effects.

't.e propr:'ed system is supposed to substantially reducecosts by accot-plishing the Government-wide reservation andticketing fur t.Lon with fewer people. Additional savings areclaimed by avoiding currently undetected ticket overcharges.The pro-o.'al also anticipates other benefits through improvedtravel ma-trigement which are not directly measurable.

The concept of a centralized Government reservationand ticketing system appears to be technically feasible.However, the estimated savings resulting from a centralizedsystem are questionable. The cost estimates were unsupportedor were beaed on unreliable information.

The proposal did not consider the practical aspects ofimplementing a centralized system. Needs of individual agen-cies were not addressed, and no provision was made foragency participation in planning for the system.

F-inally, a centralized travel system is only one ofseveral alternatives that could be considered if the Govern-ment reservation and ticketing functions were to be changed.The p-¢posal did not address other alternatives. In ourview, all possibilities should be studied before decidingwhich system best suits the Government's needs.

BACKGROUND

Advances in computer technology and automatic dataprocessing have enabled major airlines to develop reserva-tion and ticketing systems to process the increasing number

LCD-78-209(943303)

B-103315

of passengers. These reservation systems contain airlineflight schedules and seat availability information, canautomatically compute most fares between domestic locations,and can print a ticket at remote locations convenient topassengers.

The proposed system is a Government-wide automatedairline reservation and ticketing system conceptually similarto those operated by several airlines. Telephones wouldconnect the traveler anywhere in the United States with thecentral reservation center in Washington, D.C. A centralcomputer would contain flight schedules, seat availability,fares, and other information. The Government computer wouldbe connected with similar computerized systems maintainedby the airlines. Reservations would be made instantaneously,and tickets would be issued mainly through teleticketingmachines located either at Government offices convenient tothe traveler or at the airport terminal ticket counter.

To illustrate-a traveler would call a General ServicesAdministration (GSA) terminal operator in Washington, D.C.,over Government-leased telephone lines, from any point inthe United States. The traveler would specify location,destination, and flight preferences. While the traveleris still on the phone, the operator would obtain flightschedules, seat availability and other information, whichis flashed onto the terminal screen of a GSA computer. Theoperator would verbally relay the information to the traveler.The traveler would select the flight, and the operator wouldmake the reservation. Simultaneously, the computer wouldautomatically survey the traveler's flight plan for comp-liance with the Fly America Act, perform a mechanical auditto insure the correctness of flight charges, debit the ac-count of the traveler's agency, and instruct the airlinecomputer to issue the ticket through a teletype machineat a location convenient to the traveler.

IMPORTANT ASPECTS NEED MORE DEVELOPMENT

A centralized travel system is one of several possiblemethods that could be considered if the Government reserva-tion and ticketing function were to be changed. However,important aspects-of the proposal need much more developmentand consideration before a meaningful evaluation can bemade. For example:

2

B-1 03315

-- Supporting data for the economic benefits cited inthe proposal need further validation.

-- Organizational constraints and related problems needconsideration.

-- The technological feasibility of establishing acentralized travel agency needs additional study.

--The merits of all possible alternatives need con-sideration.

Projected economic benefits need validation

The proposal emphasizes the projected annual savingsof about $36 million resulting from reduced labor costsand the elimination of overcharges due to mistakes in com-puting fares. The proposal estimates current system operat-ing costs at $38 million and costs for the new system atslightly less than $2 million. A comparison of current andproposed system costs based on proposal data follows:

Estimated annual costsCurrent Proposed

Direct labor $17,200,000 $ 970,000(1,059 (72 personnel)personnel)

Indirect labor 11,800,000Overpayments to

airlines 9,000,000 -Leased automaticdata processingequipment - 960,000

Total cost $38,000,000 $1,930,000

As just mentioned, the basis for savings is the reduc-tion in the number of personnel currently processing travelreservations (direct labor) and the elimination of inter-mediate personnel (indirect labor such as secretaries) cur-rently involved in the travel process. The savings in in-direct labor would not be realized by eliminating any per-sonnel, but would represent incremental savings of timewhich could be devoted to mecre productive functions. Inaddition, overpayments in ticket fares to airlines believedto occur with current procedures would be eliminated.

3

B-103315

Direct labor

We could-not validate the current and proposed costestimates because the basic data used in the proposal tomake *the estimates was unreliable. The volume of trans-actions and the time required to process them are criticalelements to the cost estimates. Proposal estimates forboth elements are seriously deficient.

The number of transactions to be processed is a keyfactor in estimating the cost of both the current reserva-ticn and ticketing system and the centralized system. Forpurposes of this letter, a transaction is defined as makinga reservation or issuing a ticket.

The proposal estimated annual Government travel trans-actions at 1.8 million. This estimate was based on the num-ber of Government Transportation Requests (GTRs) GSA auditedduring a 1-year period. However, the number of GTRs issueddoes not represent the actual number of travel transactionsprocessed. For example, one agency we visited issued one GTRweekly for all its tickets. The 52 GTRs issued yearly re-presented about 10,868 tickets. Another agency found thatit had twice as many transactions (inquiries, reservations,and changes) as tickets issued. Also, tickets purchasedwith cash were not considered since no GTR was involved.

The proposal also assumed that 1.8 million transactionsrepresented all Government travel, including travel for theDepartment of Defense (DOD). An official of the MilitaryTraffic Management Command--DOD's single manager fortransportation--estimated that DOD alone issues 1 millionGTRs annually, representing about 4 million tickets. Soit appears that the Government issues many more ticketsthan 1.8 million each year. However, the actual volumeof Government travel in terms of tickets, transactions,and travel patterns is unknown. The basic travel informa-tion that would be needed to effectively manage traffic ona Government-wide basis is not currently collected by anagency.

Underestimating the number of transactions to be pro-cessed would understate the proposed system's operatingcosts. Since the actual number of transactions may be4 to 5 times larger than the proposal's estimate, a signi-ficant increase in personnel and equipment may be needed.

4

B-103315

The substantial increase in transactions, for example, mayrequire major computer capabilities rather than the proposedmini-computer. Also, the estimated number of personnelrequired to operate both the current and proposed systemsis probably understated, since the estimate is based on lowtransaction volume. A Defense Department official estimatedthat about 500 personnel would be needed to process justDefense travel in a centralized automated system.

The length of time required to process a transaction isalso a key factor in estimating current reservation andticketing costs. The proposal's estimated total processingtime is 1.044 hours, based on an informal, undocumentedstudy of a GSA travel office. Airline studies and otherGovernment officials, however, indicated that the totalprocessing time for manual systems is probably about 30minutes; automated systems are even faster.

The proposal assumes that the GSA travel office is re-presentative of all Government travel offices. Governmentagencies use many different travel processing systems. Someagencies avoid all travel processing by allowing each em-ployee to make travel arrangements directly with the airlines.Other agencies reduce travel processing by using co-locatedairline personnel. Some agencies have leased automated air-line reservation systems that can reduce processing time toless than 20 minutes per transaction. The wide diversityof travel processing invalidates the proposal's Government-wide extrapolation based on GSA's manual processing system.

The unreliable estimates for the number of transactionsand processing time affects the estimates for both currentand proposed system costs. The unreliability of these keyfactors precludes any meaningful evaluation of the proposal'scost estimates. Better, more precise, information must beobtained on volume and processing time before adequateestimates can be made on either current or proposed systemcosts.

Indirect labor

The proposal correctly recognizes that intermediatepersonnel such as secretaries and clerks are involved inthe travel process. Eighty-five percent of dire t labor isused as an estimate for these indirect costs.

As we understand, derivation of the 85 percent factorX's based on conversations with airline representatives.

B-103315

The proposal does not support this estimate and does notspecify how a centralized system would eliminate the needfor these intermediate personnel. Based on our observa-tions of automated reservation systems currently used inGovernment, intermediate personnel continue to be involvedin the travel process.

Savings in airline overcharges uncertain.

The proposal estimates that the Government currentlydoes not recover $9 millicn in undetected airline tic'ktovercharges. These overcharges are in addition to thosecurrently being found and recovered in GSA's routine auditof transportation payments. The overcharge estimate isbased on a study of selected travel vouchers made by aprivate company. However, the sample used in the studywas highly biased and was expected to have a high- ncidenceof fare computation errors. For example, most of the vouchersshowed travel to two or more locations on two or more airlineswhich is a type of travel particularly susceptible to farecomputation errors. The 104 sample vouchers usec. for thestudy are not typical of most Government trove', and thesample was not large enough to be statistically valid.

While the savings, if any, are uncertain f:om automaticverification of fares, we believe such a verification fea-ture has merit. It would better assure the GoiernmenL ofthe validitv of 'are charges and would further simplifycurrent audit procedures.

Equipment costs

The centralized system equipment costs were estimatedto be about $960,000. The estimates for software/hardwarewere based on informal discussions with prospective sup-pliers, and specific requirement lists were not developed.The estimates included funds for more closely defining theproblem and developing requirements.

The equipment costs were based on the pxcposa:'sestimate of travel volume. As previously discussed, thepresent proposal may considerably underestimate the numberof transactions involved. In contrast to the proposedequipment estimate of $960,000, a DOD official estimatesthat a $30 million investment in hardware/software plus$13 million to $15 million annually would be needed to operatea centralized travel system to serve only DOD personnel.

6

B-103315

Furthermore, the proposal's equipment es;timate wasonly preliminary and apparently based on a misunderstandingof the system's concept. The prospective st.pplier saidthat the estimates were for a prototype of i regionalized,not a centralized, system. Additional systems would beneeded in each GSA regional office after prototype develop-ment. Therefore, the problem would have to be more ac-cnrately defined and more specific requirements would haveto be developed before any equipment costs could be ac-curately forecasted.

Organizational constraints need attention

Before serious consi eration can be given to implement-ing a centralized automated reservation and ticketing system,we believe numerous organizational problems should be ad-dressed. Abandoning independently developed agency reserva-tion and ticketing systems in favor of a GSA centralizedsystem is a major change which requires planning, coordina-tion, and consensus. As suggested by the proposal, theOffice of Management and Budget (OMB) or GSA could compelother executive branch agencies to use a centralized system;however, a preferable approach would be to secure voluntaryparticipation and cooperation fr:m the affected agencies.

Need to overcome resistance to change

From a practical viewpoint, agencies may resist changeand oppose turning their travel service functions over toGSA. In the past agencies have individually developed sys-tems for processing travel. GSA has only provided guidanceand suggestions.

An example of resistance to change concerns the re-luctance of some agencies to adopt automatic airline ticketpayment plans and teleticketing procedures, although thesetechniques have been shown to be cost effective. Moreover,because ticketing and reservation systems are important toan agency's operations, assurance must be given that acentralized system would perform effectively without inter-ruption. Interagency coordination would provide such assur-ance and ease the resistance to implementing a new system.

Agency travel needs must be identified

Agencies have independently developed travel proceduresto meet their needs. These needs differ and would have to

7

B-103315

be determined and analyzed. For example, DOD travel representsover half of the Government's total travel requirements. DODtravel needs are different from those of civilian agencies,as evidenced by the provisions of DOD travel regulations.Any proposed system would have to be coordinated with DOD

and other agencies. By doing this, potential interfaceproblems would be identified and resolved. We believe thatthe varying travel needs can only be identified through wideparticipation of all agencies in the planning and designingof a new system.

The current proposal has not, in our opinion, givenadequate consideration to agencies' specific needs.

Cooperation needed to solve interfaceproblems concerning agency accountingsystems

The proposal does not discuss how airlines are reim-bursed. Some agency accounting systems restrict the wayairlines are paid. For example, one agency that we reviewed

has an automated centralized system for the Washington, D.C.,

area; however, budgeting and accounting procedures in thisagency require each bureau within the agency to individually

make payments to the airline. Differences in agency account-ing systems and payment procedures indicate the need for acooperative agency effort to consider these problems.

Proposal system technically feasible

The proposed system appears technically feasible be-cause of its similarity to systems used by major airlines

and leased to Government agencies and private corporations.Most officials with whom we discussed the proposal acknow-ledged the technical fiasibility of such a system.

Using computers to provide real-time rese:vation andticketing services is clearly feasible as demonstrated bythe airline systems. Visual display units, telephone com-munication with travelers, ald the ability to instantlyissue tickets via teletype machines have been used suc-cessfully throughout the country. In addition, the pro-posal offers potential for more meanlngful travel manage-ment reports combined with better accounting and payingprocedures.

8

B-103315

Automated systems have alloied reservation and ticketingpersonnel to increase productivity. The problem of handLingincreased numbers of passengers and the ability to rapidlyprocess transactions have been made possible by adoptingthese systems. However, a number of technical questionsremain to be resolved before this proposal could be imple-mented in the Government. For example, equipment require-ments to cope with emergencies and demand surges must bedetermined.

Alternative methods not considered

The proposal assumes that a single Government-widereservation and ticketing system is the only viable alter-native to present methods. However, there are a number ofpotential alternatives that could be considered. Some ofthese alternatives include:

-Expanding use of airline-developed automated reserva-tion systems. Several Government agencies and privatesector corporations use airline-develuped systemswhich essentially do everything that the proposedsystem envisions only on a smaller scale.

--Expanding agency use of scheduled airline ticketoffices, whic a e staffed by airline reservationpersonkel an6 currently located in many DOD andcivilian installa ions.

-=Expanding agency use of teleticketing machines andautomatic payment provisions. This could providesavings in reservation and ticketing. GSA has recom-mended these techniques for several years, but notall agencies have adopted them.

---Establishing multiagency, regional, or metropolitanarea automated reservation and ticketing systemsSimilar to the proposed system, but on a smallerscale. The increased capacity of the airline systemsoffers the potential for several agencies to use asingle system and share the costs. DOD officialsat the Military Traffic Management Command said thatrecognizing the advantages of airline-dev-loped sys-tems has led to the Command considering a DOD-wideticketing and reservation system.

11

B-103315

--Permitting Government travelers to independentlymake travel arrangements without agency assistance,decreasing the direct involvement of Government per-sonnel.

In our view, the beneit-s and costs of ail feasiblealternatives should be considered before recommending anyparticular method.

COST SHARING NEEDS CONSIDERATION

An additional underlying consideration in evaluatingalternative methods is the extent to which airlines or theGovernment should bear the cost of this activity. Thesecosts are normally considered an airline responsibility,and they are part of the cost base used by the CivilAeronautics Board to establish fares. In providing reserva-tion and ticketing services, the Government seeks improvedadministrative control over travel and convenience to Govern-ment travelers. On the other hand, the Government relievesthe airlines of considerable costs that they would ctherwiseincur. Also, Government airline tickets at full fare areusually paid prom'ify, avoiding any lessened tavenue tocarriers from rdying travel agent, bank, anl entertainmentcaLd commissions. Consequently, the relative costs to beassumed by the Government and the airlines would requireextensive consideration.

PRIVACY ACT IMPLICATIONS NOT CINSIDERED

Since the Government-wide reservation and ticketingsystem would contain personal information--naiues of individ-uals and locations to which the individuals travel--thesystem would fall within the purview of the Privacy Act of1974. The Privacy Act was enacted as a safeguard againstthe threat posed to individual privacy by the Federal Govern-ment's collection and use of personal information, and itimposes certain requirements upon agencies. The adminis-trative costs for implementing these requirements would addto the cost of the proposed system.

CONCLUDING COMMENTS

The results of our evaluation were discussed separatelywith GSA officials and with the individual who submitted theproposal. No major disagreement was expressed about thefindings presented in this report.

10

Ba103315

The individual who made the proposal acknowledged thelimitations of his estimates. He said that the proposalwas the best that could be developed, considering the dif-ficulties in obtaining travel information arid the limitedtime and resources he was allowed to devote to the study.Be continued to support the centralized concept.

Notwithstanding the limitations of the present proposal,this individual has made a valuable contribution to ongoingefforts to improve Government travel management by focusingattention on the potential of a centralized travel system.GSA started an internal study on several reservation andticketing alternatives, including a centralized system.

We believe that when the study is completed, OMB andGSA should jointly consider what actions would be approp-riate to see that Government- departments and agencies fullyconsider the possible benefits and savings from the Govern-ment ticketing and reservation system. Also, we suggestthat OMB and GSA examine the extent to which current costsof the Gc ernment ticketing and rez. ration function areequitab.l divided between the GovernmeD; and the airlines.

You., ,ffice requested that we make no further distri-bution o. this report until you have had an opportunity toreview i- and to prepare a response to the individual whosubmitted the proposal. Your office also indicated thatyou may request OMB to study some alternative concept.After you have responded to the above individual, we willsend copies to interested parties and make copies availableto others upon request.

Sincerely yours,

P. J. ShaferDirector