Chapter 8 Moving Toward a High-Skill Economy

Chapter 8

Moving Toward a High-Skill Economy:Computer Applications and Work

Organization in the Services

CONTENTS

PageSummary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ........................259Service Jobs and Manufacturing Jobs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .262Using Computer and Communications Systems . . . . . . . . . . . . . ............263

Business Applications of Computer Systems.. . . . . . . . . . . . . . . . . . . . . ....265Motivations: Cost Control and Strategy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .267Integration: Technical and Organizational Dimensions . . . . . . ...........269

Systems Design and People’s Skills . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .272Integrated Production Systems: The New Model. . ......................278Impacts on Competitive Ability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. ....282Concluding Remarks. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ............285Appendix 8A: Restructuring and Work Reorganization at MetroBank .. ...288

BoxesBox

Z. Claims Processing in the Insurance Industry . . .AA. Technological Advances in Computing SystemsBB. Tasks, Jobs, and Skills . . . . . . . . . . . . . . . . . . . . . .

FiguresFigure No.48. Intra-Corporate Information Flows . . . . . . . . . . . .49. Characteristics of Firms and Industries . . . . . . . .

TablesTable No.

Page● . . . . . . ... 264● . . . . . . . . . 267. . . . . . . . . 2 7 6

Page. . . . . . . . . 270. . . . . . . 277

Page41. Examples of Strategic Applications of Data Processing and

Communications Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ....0.26942. IBM’s Intra-Corporate Computer Networks . . . . . . . . . . . . . . . . . . . . . . . . .27043. Patterns of Computer Use by Occupation. . . . . ......................27644. Typical Changes in Cost Structure for New-Model Organizations .. ....284

Chapter 8

Moving Toward a High Skill Economy:Computer Applications and Work

Organization in the Services

The preceding chapter outlined the patternsemerging in the U.S. labor force over the past15 years—a period during which manufactur-ing employment stagnated and began to declinewhile job opportunities in the services con-tinued to expand. Some of the jobs created havebeen low-skill, low-wage positions in the ter-tiary services; others have been knowledge-based, with far better prospects for upward mo-bility and job security. Where possible, chap-ter 7 examined these trends quantitatively, interms of numbers of people employed, theirages, educational backgrounds, earnings,

This chapter is descriptive rather than quan-titative. In places it is speculative. The chapterexplores work organization and computer ap-plications in the services, asking: How does in-ternational competitiveness depend on skillsand knowledge? How do companies make useof computer-based systems to enhance or to re-place the capabilities of their employees? With74 million people in the Nation’s service workforce, and some 38 million in the knowledge-based sectors, generalizations remain open toquestion, But conclusions do follow. To get andhold good jobs in the services, Americans willhave to bring better knowledge and skills to theworkplace. Public education will have to re-spond to new demands. Companies will haveto provide better training, and do so on a con-tinuing basis. In the knowledge-based services,employees at all levels will find themselves tak-ing on more responsibility. And, while largenumbers of relatively unskilled (and low-pay-ing) service jobs will remain in the U.S. econ-omy, many of these jobs will depend on pros-perity created in part by knowledge-basedservices.

More often than not, technology in the serv-ices means applications of computer and com-munications systems, The knowledge-basedservices are in the midst of a transition fromlarge and expensive mainframe computers,tended by experts, to distributed computing,used by everyone. The personal computer isonly the most obvious sign of this transition,Yet before companies and their employees cancome to grips with today’s technology, newwaves of hardware and software will arrive—making it much easier, for example, to link PCswith powerful mainframes, helping create far-flung integrated networks. People and organi-zations will have a hard time keeping up, muchless planning for the future.

In many U.S. service firms, proprietary tech-nologies—including computer applications—have become integral elements in competitivestrategy. Isolating the contributions of propri-etary know-how tends to be a good deal moredifficult for service products than for tangiblegoods. Manufactured products can be inspectedand tested, performance evaluated by objectivemeasures, making it easier to judge relative con-tributions of proprietary technology, and—much the same thing—human capital. Chap-ter 3 illustrated some of these complexities forbanking, chapter 5 for computer software; soft-ware is much the easier case because computerprograms can be compared on a price/perform-ance basis far more readily than banking serv-ices. But when the question becomes: Preciselywhy is one program better than another? an-swers can be hard to find. In the end, proprie-tary technology is a matter of judgment and ex-perience more than well-codified knowledge.Individual expertise makes a difference; so do

259

260 . International Competition in Services

group and social skills. Both the people in acompany and their tools matter. When a com-pany decides to buy a million-dollar softwarepackage from a vendor, its employees must se-lect the right package; typically they will alsocontribute to design modifications that tailorit to the company’s needs.

No one can specify with any precision howhuman capital affects competitiveness, al-though no one denies its importance. Much ofthis chapter focuses on applications of com-puters, because, for many years to come, thisset of technologies will have enormous impactson competitiveness in the services, and on thejobs and skills of people who work in serviceindustries. At the micro level, work organiza-tion—how companies design tasks for individ-uals (e. g., data entry), and combine these tasksinto groups (e.g., for processing financial trans-actions) —strongly influences the cost and qual-ity of service. Some companies have sought todevelop flexible systems for the delivery of serv-ices (and goods) —systems highly responsive tomarket demands (e.g., portfolios of mutualfunds, among which customers can switch), aswell as strategic applications of computingpower such as on-line customer ordering, orcorporate cash management systems. But cheapcomputing power also creates opportunities forrigid, mechanistic forms of automation (checkprocessing in many banks, directory assistancefrom the telephone company). Sometimes im-proved competitiveness calls for flexibility,sometimes for cutting costs through rigid andtightly controlled forms of work organization.Obviously, a vast middle ground separates thesetwo poles.

Over the longer term, strategic applicationsleading to new products, different services, andexpanding markets will have greater impactson international competitive standing than cost-cutting applications of computers. Some ofthese strategic applications will emerge at themicro level of individuals, work groups, anddepartments; they can be viewed in a workorganization framework. Examples includeCitibank’s replacement of centralized trans-action processing by a decentralized, product-oriented system—described later in the chap-

ter. Other strategic applications must be viewedin organization-wide terms—often at the macrolevel of multinational integration.

Despite a shortage of concrete informationon how computer applications and computerintegration affect the competitive ability of par-ticular companies, at either micro or macrolevels, OTA’s analysis suggests differences inapproach internationally, which appear totranslate into differences in competi t ivestrength. At the micro level of work organi-zation—integration of people and machines,rather than integration of dispersed corporateoperations—many foreign firms do as well orbetter than their American counterparts. At themacro level of international integration, U. S.-based firms seem to be well ahead of their for-eign competitors in the ability to link and co-ordinate the activities of divisions and affili-ates thousands of miles apart.

American companies may spend more moneyon computers, but some foreign firms—partic-ularly in Japan—use the money they spend moreeffectively. They scrutinize investment deci-sions for hardware and software more closely,and at more levels in the organization, investmore heavily in the training of their people, andoperate highly developed systems for maximiz-ing individual contributions to corporate goals.(These differences seem more evident in man-ufacturing than the services; Japanese serviceindustries, with exceptions such as telecommu-nications and banking, appear relatively un-developed compared to American industries—although this may simply be because analystsin the West have not focused as much atten-tion on them.) Large Japanese corporations lagbehind their American counterparts in the rawcapability of their installed computer and com-munications systems, but use what they haveat least as effectively. Moreover, concerted ef-forts in Japan to move towardan‘‘informationeconomy” (chs. 5 and 9) suggest that Japanesecompanies may begin to catch up in computerutilization during the 1990s,

At both micro and macro levels, current po-sitions matter less than decisions being madetoday. These decisions—how to use computer

Ch. 8—Moving Toward a High-Ski/l Economy: Computer Applications and Work Organization in the Services ● 261—

and telecommunication technologies (whichwill depend in part on the capabilities of thepeople a company can hire, and on the train-ing it chooses to provide) —will influence com-petitiveness 5, 10, 25 years from now. Managersface difficult decisions, A great deal of tech-nology exists; many of the possible applicationsremain largely unexplored. As noted below, no-body knows the capabilities that artificial in-telligence (AI) may bring to the workplace inthe 1990s. The specialists who develop the sys-tems tend to be interested in technology for thesake of technology, rather than in appropriateapplications. Some companies plunge ahead,making investments and reorganizing work inthe hope of establishing a competitive advan-tage, even though they may have no more thana hazy idea of the outcomes to be expected.Others move more slowly, sticking to what theyknow and understand. Sometimes one ap-proach will prove right, sometimes the other,

Two examples—one at the micro level, oneat the macro—illustrate the uncertainties.Today, many companies must ask whether AI,touted for 25 years, is finally ready to enter theworkplace on a large scale. Expert systems in-tended to supplement people’s skills have be-come one of the favored near-term applications,What should the prospective user—bank, insur-ance company, medical clinic—do? Invest itsmoney, time, and effort now? Or continue towait, at the risk of losing out to competitors whoget a head start? At the macro level, the well-publicized strains between General Motors andits EDS subsidiary illustrate another set of prob-lems, GM purchased EDS in large measure toget help in implementing its strategy of mul-tinational integration (ch. 5). Since the acqui-sition, GM’s management has come in forample criticism, Some of it is no doubt justi-fied, but the fact is that the task of puttingtogether and learning to use a worldwide com-puter and communications network is enor-mously complex, the territory largely uncharted,mistakes inevitable.

Despite the confusion and uncertainty suchexamples hint at, the outlines of a new modelof computer-integrated production have begunto emerge. The model fits some companies in

both the services and manufacturing, but so farprobably only a few thousand American firmsin total. The common characteristic of thesenew-model firms is their combination of effi-ciency and flexibility. In the past, flexibility—the ability to respond to changing market con-ditions, to alter the firm’s output or way of do-ing business (because of shifting consumertastes, ups and downs in the business cycle,competitive pressures, new technological op-portunities)—generally came at the price of effi-ciency. Flexibility meant labor-intensive oper-ations. People are flexible, they can adapt.Machines, in the past, could not, The trade-offwas a simple one: either a flexible organization,labor-intensive and relatively low in produc-tivity, or an inflexible, mechanized productionsystem, Mechanization brought higher produc-tivity but also higher costs for adapting tochange. What the computer brings is the po-tential (not always realized) for both produc-tivity and flexibility,

Achieving both flexibility and efficiencyplaces new demands on the labor force. Em-ployees at all levels in new-model organizationsmust take on greater responsibilities. Integratedsystems will put them in touch with more peo-ple, both inside the firm (colleagues in workgroups, people in other departments) and out-side (customers with problems or inquiries).Some kinds of work, in some companies, willbe more fluid, less predictable. Some jobs willbe upskilled; the people that fill them will needboth broader and deeper skills—problem-solv-ing and reasoning, social skills. Companies willrely on employees with these skills to compete.People will need these skills to get good jobsand keep them, to advance. New applicationsof computer and communication technologiesmean new demands on the public educationsystem; they will also mean new kinds of cor-porate training programs. At the same time,other jobs will be deskilled: the computer willmake them simpler, more routine, less demand-ing, less interesting. Stratification in skills goeshand in hand with the stratification in wagelevels and mobility prospects discussed in chap-ter 7. To keep job ladders and mobility chan-nels open, the U.S. economy will need to con-

262 ● International Competition in Services

tinue creating large numbers of jobs at thehigh-skill end of the spectrum; policy makerswill need to make sure the United States cre-ates education and training systems that canprepare people for these jobs.

This chapter, then, suggests that a labor forcerich in the skills needed for integrated produc-tion systems—in which applications of com-puter technologies enhance people’s skills andcontribute to organizational flexibility-will aidAmerican firms in international competition,contributing to job creation and the Nation’sstandard of living. Indeed, to compete effec-tively with economies having lower wages and

living standards, the United States has literallyno option but to create greater numbers of jobsin which computer and telecommunicationssystems enhance people’s skills, helping themproduce higher value-added services and goods.This implies better education and training,especially for new entrants to the labor marketand younger employees; a high-wage economycan only remain competitive through continu-ing investment in human capital. The alterna-tive? Continued loss of ground to low-wageeconomies, followed by declining wages andlower living standards in the United States.

SERVICE JOBS AND MANUFACTURING JOBS

Preparing hamburgers in a fast-food restau-rant has a good deal in common with assem-bly work in an automobile factory; the job ofa white-collar clerk feeding forms into an opti-cal character reader resembles that of a blue--collar employee tending a punch press in ametal stamping plant, or a kitchen worker load-ing food onto trays in a hospital, As such ex-amples imply, the nature of a job may dependas much on work organization as on whetherthe job falls nominally on the service side orthe manufacturing side of the economy. Indeed,viewed as encompassing skills, expertise, know-how, and work organization, technology be-comes a major competitive weapon even inmany of the tertiary services. The most casualobserver sees the differences among fast-foodrestaurants, Each faces the same fundamentalproblem: defining a menu; managing a high-turnover labor force with limited skills and lit-tle job experience; dealing with a work load thatfluctuates greatly during the day. How to or-ganize production to give customers what theywant, when they want it, while minimizingoperating costs? Each chain has reached itsown solution—a matter of proprietary technol-ogy in the sense of firm-specific knowledge andexpertise, standardized procedures, trainingmanuals. Some look more like production lines;others emphasize customer choice,

Differences between the services and manu-facturing run deeper when it comes to jobs forprofessionals, In manufacturing firms, most ofthese people fill managerial and staff positions.They may supervise production in manufac-turing, but in the services many professionalswill be an integral part of the process; lawyers,surgeons, and teachers produce service outputswith their own minds and hands. In an account-ing firm, professionals carry much of the re-sponsibility for direct production; the accoun-tants employed by a stamping company are nomore part of the production process than thesupervisors.

Despite such contrasts, numerous similari-ties emerge when comparing knowledge-basedservice jobs with high-technology manufactur-ing—the most obvious being applications of dig-ital electronics to rationalize production, sup-port managerial decisionmaking, and aid indesign, development, and marketing, If sometertiary service firms depend heavily on pro-prietary technology, almost all those in the in-termediate or knowledge-based services do. Allfinancial service firms of any size, for instance,have had to develop the internal expertiseneeded to implement complex computer appli-cations. Hartford Insurance Group employswell over a thousand programmers. Airlines

Ch. 8—Moving Toward a High-Ski// Economy: Computer Applications and Work Organization in the Services . 263

that have developed their own reservation sys-tems have reaped competitive advantages; withprice competition holding down differences infares, quick and easy bookings for customerswho telephone directly, a well-established net-work of relationships with travel agents, andresponsive service for major corporate custom-ers count for a good deal. Note that an effec-tive reservation system depends not only oncomputer hardware and software, but also onthe airline’s employees—their training, respon-siveness, and commitment to their employer’sgoals. Everything else the same, the airline thatmakes customer service a genuine objective,and conveys that successfully to its employees,should be able to fill more seats than its com-petitors,

In many industries, knowledge and skills ulti-mately determine international competitive po-sition; there is plenty of truth in the saying thatinvestment bankers live off their wits. Regard-

less of industry, the capabilities that peoplebring to the workplace affect productivity andcompetitiveness: through technical skills, thoseof the software engineer or the computer repair-man, loan officer or insurance underwriter;through managerial and administrative skills,those of the buyer in a department store, thebranch manager in a bank, the foreman in acopper mine. The work may range from find-ing investment capital at the lowest costs, toplanning new products, to lobbying govern-ment agencies, Sometimes the knowledge thatmatters is well-codified (as in the computer-aided engineering methods used to design ahydroelectric power station), sometimes tacit(as in the experience-based judgment that bankemployees bring to the arrangement of financ-ing for that power station). In all cases, the tech-nology is developed by people, embodied inpeople’s skills, transferred and diffused bypeople,

USING COMPUTER AND COMMUNICATIONS SYSTEMS

In industries like financial services, somecomputer applications simply automate por-tions of an existing production process. Otherapplications enable firms to create new serv-ice products, or provide existing services innew ways. Prentice-Hall now supplies its taxservice, formerly delivered in loose-leaf form,on-line to law and accounting firms, as well asto corporate tax and legal departments.

New products can lead to new jobs, Whetherthey door don’t, they typically mean new skillsand new learning for the existing work force.Appendix 8A, at the end of the chapter—a casestudy of restructuring at MetroBank (a fiction-alized name) —illustrates. MetroBank’s strategyrequired that customer service representatives(CSRs) actively sell a new line of products (e.g.,credit cards, individual retirement accounts),In the process, the CSRs had to learn to usea redesigned and expanded computer systemgiving on-line access to customer accounts.Managers whose roles had been largely admin-istrative had to learn to coach the CSRs, as well

as to market the bank’s services, includingloans, to business customers. Moving into thesenew roles proved difficult, and sometimes pain-ful, for both sets of employees. MetroBank’s sys-tems division, meanwhile, carried out its re-design of the computer network with littleattempt to understand the CSR’s working situ-ation and needs, aggravating the difficulties.This bank’s experience is not unusual; similarstories can be heard in other financial servicesfirms and in other industries.

In some applications, the computer followsthe same rules and procedures as the peopleit supplements or replaces: knowledge—oncethe monopoly of people with skills and experi-ence—becomes embedded in the system, BoxZ, which traces the evolution of automatedclaims processing in the insurance industry,describes how companies have put computersto work doing what people once did, faster andcheaper.

As box Z shows, work has changed radicallyfor both clerical employees and claims adjus-

264 ● /international Competition in Services

Box Z.—Claims Processing in the Insurance Industry1

Insurance companies began automating in the1960s, installing computer systems for batchprocessing of high-volume but relatively routineclaims—group health plans, automobile cover-age. Labor intensity remained high through themiddle 1970s, with clerks continuing to checkand double-check the process at many points.These partially automated systems functioned asoutlined below, using workers’ compensation asan example:

1.

2.

3.

4*

Claims went first to an adjustor, who wouldverify the loss and the identity of the insured.The adjustor then filled out an instructionsheet, by hand, got it approved by a super-visor, and sent the sheet to a typist who pre-pared a data entry coding form.Each day, a batch of these forms went to thecompany’s data processing center, where aclerk would pull a copy of the policy to checkthe coverage levels, then complete a differ-ent form and send it on to the keypunchdepartment. If the policy were missing fromthe file—often the case—the clerk had towrite to another office for a duplicate of theoriginal rating sheets. A claims coder wouldthen recode the policy before the claim couldbe processed and sent for keypunching.After keypunching, the computer took over,printing checks to be mailed to claimantsand recording payments. Processing tookplace in batches at night. Clerks continuedto log each payment by hand. These logswere reconciled with the computer recordsonce a year.

This process may seem complex, but others weremore so; before paying a fire insurance or au-tomobile claim, for example, an examiner wouldhave to inspect the damage. Often, negotiationswith other insurance carriers or with repair com-panies would follow.

Work reorganization in this industry has beendriven by domestic competitive pressures (thereis little international competition), with compa-nies striving to cut costs and increase produc-tivity in order to improve profitability and mar-ket share. Around the middle of the 1970s,

IAdapted from “Draft Report: Insurance,” prepared by B. Baranunder contract for the OTA assessment, Technology and the Amer-ican Economic Transition, pp. 49-54.

insurance companies began to invest in on-linesystems. Here, two alternative patterns of workorganization have emerged, one with claims ad-justors as end users, the other with most tasksperformed by clerical employees.

●

●

Where adjustors use the system themselves,clerks first screen and sort incoming claims.The adjustors investigate, authorize, andprint settlement checks at their terminals,avoiding most of the intermediate steps ofthe older batch processing procedures.With clerical workers as end users, the soft-ware must contain decision rules that candispose of the majority of claims. When aclerk at a terminal runs into a case that thesystem can’t handle, she or he calls on anadjustor for help.

Productivity has grown enormously with bothapproaches, which differ in the ratio of morehighly skilled adjustors to clerical workers, andin the intelligence built into the system.

Today, vendors market dedicated systems forhealth care claims that function as follows:

1.

2.

3.

4,

Incoming claims go directly to a clerk or as-sessor, who checks the form for complete-ness and assesses the claim (for coverage,contractual limits, reasonable and custom-ary limits).The operator calls up files with personal in-formation on the insured [and may amendthem), while the system creates windowsthat display allowable payments and ex-planatory codes (e.g., for physician charges,laboratory tests, drugs).Using other windows, the clerk or adjustercan take account of co-insurance, cata-strophic clauses, deductibles, and yearly orlifetime maximums,A final window displays a draft paymentform for the operator to verify or modify, andapprove.

These systems not only issue checks automat-ically; some can prepare form letters withupwards of 2,000 variations. The system willautomatically generate accounting and manage-ment reports that help the company predict claimfrequencies and estimate its loss ratios. It can alsotrack employee productivity. With insurancecompanies continuing to extend the capabilitiesof these systems, clerical workers (where they are

Ch. 8—Moving Toward a High-Skill Economy: Computer Applications and Work organization in the Services ● 265

the operators) can handle two-thirds or more ofall claims with no need to call for help from anadjustor.

The new systems eliminate coding, keypunch-ing, and manual verification. With most of therepetitive and redundant tasks gone, opportuni-ties for mistakes are fewer. So are opportunitiesto catch them. While some checks on accuracycan be built into the software, employees—whether clerks or adjustors—must take most of

tors in the insurance industry. Clerical jobs havegrown more demanding, with people previ-ously viewed as unskilled asked to take on manyof the responsibilities of adjustors and asses-sors, While clerical jobs have been upskilled,pay has not changed much. The work of theremaining adjustors has also been upskilled,They get fewer routine cases; more of theirworkload consists of claims that the computer’sbuilt-in decision rules cannot handle,

In other commonly found patterns, it is themiddle levels of knowledge and skill that be-come part of the system—leaving people with,say, data entry jobs at one extreme, and highlyskilled tasks beyond the computer’s capabilitiesat the other. In effect, work reorganizationdeskills some jobs while upskilling others. Even-tually, many of the jobs like data entry will alsodisappear (because operators working at on-lineterminals normally do this themselves).

As digital systems become still more perva-sive (in many industries, their power has barelybeen felt), their influence on the organizationof work still greater, both people and organi-zations will have to learn and adapt. This meansdesigning applications that are well-suited tothe skills, aptitudes, and motivations of thefirm’s employees, integrating their skills andabilities with those of the system—no easy task.Suppose, for example, that a multinational bankdecides to invest in a computer network forlinking its branches and subsidiaries. Viewedas a straightforward application of availabletechnology, the critical skills lie in the designof the system: choice of equipment; when touse leased lines and when to rely on the public

the responsibility for the correctness and com-pleteness of their own work.

The next steps? A number of large automobileinsurers have begun to give adjustors in the fieldportable PCs. By dialing into computers at abranch or main office, adjustors get the namesand addresses of claimants to be contacted. Theycan print checks with their PCs. With almost allthe work done remotely, there are no backupsexcept those that can be built into the computersystem.

infrastructure; software for running the net-work, But viewed as a means for the bank tominimize financial risks and maximize prof-its, the problem is to develop a system with in-stalled performance (as opposed to designspecifications) that will meet the needs andcomplement the abilities of the bank’s world-wide staff. Such a view helps clarify the difficul-ties involved: few if any of the bank’s operatingemployees will understand computer networks;the system designers will not understand bank-ing. Still, if the bank manages this task well,it may be able to capitalize on fleeting differ-ences internationally of l/32nd of a percent-age point in interest rates. If it manages the taskpoorly, the network might be close to useless,and require extensive redesign.

Business Applications of Computer Systems

Over much of the 30-year history of computeruse within business organizations, companieshave simply automated existing tasks. Insur-ance companies began with batch processingof claims, as described in box Z. Banks learnedto process checks more cheaply, helping themkeep up with rapidly expanding transactionvolumes. Businesses of all kinds automated rou-tine functions such as payrolls to cut costs. Ana-lytical applications such as computer-aided de-cision support for financial risk analysis camelater,

Today, computing power is so cheap that itsapplications have become part of everydayworking life for millions of Americans, Newapplications can be tailored to the requirements

266 • IInternational Competition in Services

of upper managers and executives (work sta-tions for the corporate treasurer). Companiescan use computers to regulate and monitor rou-tine production (as in the well-publicized caseswhere telephone operators find their work over-seen by call-monitoring systems).1 Some peo-ple find themselves with jobs that are more chal-lenging and perhaps more rewarding; othersfind themselves part of a mechanized systemlittle different from a 1930s-era assembly lineexcept for the computer at its heart.

In the early years, as computers began toproliferate, corporate data processing (DP) de-partments with professional staffs took mostof the responsibility for selecting hardware andsoftware, particularly in the larger firms thatpioneered back-office applications in bankingand insurance. Smaller businesses began pur-chasing systems during the latter part of the1960s, as prices dropped.2 Applications broad-ened well beyond the accounting and records-keeping packages that had been on the shoppinglists of most first-time business customers, Dig-ital equipment also began turning up on the fac-tory floor for industrial process control.

The centralization of the early years startedto wane during the late 1970s—a result of dis-tributed processing, friendlier systems, and, afew years later, the spread of personal com-puters. With perhaps 8 million PCs in use inAmerican businesses—half as many as havebeen bought for home use, and a penetrationthat remains below 15 percent—massive newwaves of expansion and technological changelie ahead (box AA).

Business purchases of PCs, along with con-tinued progress in networks and distributedcomputing, have helped destroy the monopo-lies that centralized DP departments once en-joyed in American corporations. With largefirms moving toward decentralized informa-

ISee, for example, M.W. Miller, ‘‘Computers Keep Eye OnWorkers and See If They Perform Well,” Wall Street journal,June 3, 1985, p. 1; W. Serrin, “More Workers’ Terminals AreStaring Back, ” New York Times, May 14, 1986, p. B8,

‘See, for example, Appendix C, “Computers: A Machine forSmaller Businesses,” International Competitiveness in Elec-

tronics [Washington, DC: Office of Technology Assessment, No-vember 1983), pp. 531-535.

tion utilities, a few have even converted theirDP centers to employee training facilities. Al-though some companies continue to write theirown software, standardized applications pack-ages have taken over much of the corporatemarket in the United States (ch. 5). When ven-dors like McCormack & Dodge and MSA sellintegrated software for accounting (generalledger, accounts payable and receivable, fixedassets, personnel and payroll, purchasing, in-ventory management), there is no longer muchpoint in a firm putting its own programs to-gether.

The need today? Software and systems appli-cations for strategic purposes. Most firms thatoperate on an international scale have alreadyachieved many of the savings possible throughautomation of existing functions. The nextwave of applications will help them delivergoods and services to customers more effec-tively. American Hospital Supply, to take oneexample, has linked its computers with thoseof hospitals and clinics, which can now placetheir orders electronically. In such cases—i.e.,if a firm can establish a competitive edge witha unique software package—internal develop-ment may still make senses

%ome companies have also chosen to market software origi-nally developed for their own use, as a new line of business orto recoup some of their development costs. Accounting firmslike Arthur Anderson and Peat, Marwick, Mitchell have beenmarketing software packages to their customers for years, as havea number of investment banks—E. D. Myers, “Big Eight V.ADAPSO?” Datamation, Jan. 1, 1986, p. 32; P Hodges, “Do theBig Eight Add Up?” Datamation, Feb. 15, 1987, p. 63.

Salomon Brothers, for another example, offers software foreconomic forecasting and equity screening; originally developedto support their in-house investment management operations,outside sales have been directed at institutional investmentmanagers. Manufacturing firms including Westinghouse, Stand-ard Oil, Republic Steel, and Boeing have established subsidiariesfor marketing soft ware (or computing services) to other compa-nies. Boeing Computer Services sells time on its Cray X-MP super-computer, A telecommunications link between Boeing’s officesin England and the United States gives engineers at Britain’sNational Nuclear Corp. access to a simulation program origi-nally developed by Lawrence Livermore Laboratory for the U.S.Department of Energy. The program predicts the consequencesof failures in the cooling system of a nuclear powerplant. See“Boeing Draws on Its Years of Experience, ” Financial Times,Oct. 14, 1985, p. 12.

Ch. 8—Moving Toward a High-Ski// Economy: Computer Applications and Work Organization in the Services ● 267

Box AA.-Technological Advances in Computing Systems

In the services, computer and telecommunica-tions systems do two fundamental kinds ofthings:

1.

2.

provide error-free management, manipula-tion, and transmittal of almost unimaginablylarge volumes of data and information (theprimary sources of error will normally be atthe input end, except for software bugs andsystem design flaws); andsolve mathematical (and logical) problemspreviously intractable, leading to new ana-lytical tools (expert systems for decision sup-port) and real-time control models for pro-duction processes (optimization of messagetraffic in a telecommunications network).

Over the medium term of 5 to 10 years:• Businesses will continue to invest heavily in

distributed systems and dispersed comput-ing power, with cheap machine intelligenceand inexpensive mass storage available inmany locations (through local and wide areanetworks, LANs and WANs—ch. 5). * Half or

* Of 40 multinationals surveyed for OTA, more than 80 percent hadintroduced or were planning to install LANs, with office automationthe most common initial application—’’Data Processing in Multi-national Corporations,*’ draft prepared for OTA under contract No.533-6410 by Mackintosh International Ltd. This survey, which formsthe basis for portions of the analysis elsewhere in the chapter, cov-ered multinational corporations (MNCs) with headquarters in theUnited States, Europe, and Japan (16 American, 16 European, and8 Japanese). Seventeen companies were primarily service suppliers;the rest did most of their business in manufacturing.

The survey found that many managers, somewhat at sea with thepossibilities of computer-based systems, feel there is too much newtechnology to successfully understand and utilize. Not only domanagers and lower level emp[oyees who feel intimidated or threat-ened by the new technologies resist learning about them, but manyexecutives seem confused over the economic benefits. The MNCsmost comfortable in this environment tended to be those already quitefamiliar with the new technologies, either because they produce themor because they have lengthy experience as users (computer manu-facturers, aerospace companies, large financial services firms).

On expert systems in financial services, below, see ch. 3; on ISDN(Integrated Services Digital Networks), see chs. 5 and 9.

Motivations: Cost Control and Strategy

Multinationals expand and improve theircomputer and communications networks to cutcosts and/or pursue new business strategies.Their expectations may or may not be met; inmany cases, system performance fails to liveup to expectations—or, putting it another way,

more of major U.S. financial services firmsexpect to have expert systems installed onat least an experimental basis by 1990. (Oneresult will be to begin a new wave of deskill-ing in banking.)Companies will use these distributed com-puting systems for manipulation and com-munication of text and graphics, as well asquantitative data, the system becoming atelecommunications network and informa-tion storage device as much as a calculatingmachine. This evolutionary change will cul-minate in the widespread availability ofISDN services.The growing range of inexpensive, off-the-shelf software for small machines will enablepeople without specialized training to usecomputers for routine applications that gowell beyond the word processing and book-keeping common today. Examples includedesktop publishing and much more powerfulgraphics packages. While falling prices forhardware lie behind the explosive growth incomputer applications in business and in-dustry, it is software that determines whatcomputers can ultimately do.

The early expert systems for banking will be ex-pensive. Standardized software packages for PCsare cheap. ISDN will be a massive and costly un-dertaking. The point is simply to suggest some-thing of the directions and scope of technicalchange. Within the next few years, still less costlymass storage on optical discs will provide a ma-jor boost for business applications. Somewhatfurther ahead, corporate users should be able tobegin making extensive use of AI, and, perhaps,natural language processing. Where very largedata bases and fourth-generation languages com-bine to overload the largest current business-oriented systems, some companies may turn tosupercomputers.

the impacts on cost structure and competitiveability differ from those anticipated. Often, nomore than a hazy idea of future benefits willbe possible in advance, This was probably thecase with IBM’s venture into a unified world-wide engineering and manufacturing databaseduring the early 1970s—an undertaking that evi-dently proved far more time-consuming, costly,

268 • International Competition in Services

and painful than the company anticipated. Withthe focus shifting toward providing new serv-ices and entering new markets, risks will goup because failure may endanger a company’sstrategy as well as its cost structure, But to theextent that strategic applications work out asplanned, the company may benefit in indirectways, hard to capture with conventional ac-counting measures. For instance, a firm maybe able to create new forms of customer loyalty.

In the United States, particularly in contextssuch as computer-aided manufacturing, invest-ment decisions have frequently been criticizedas short-sighted, most commonly on groundsthat managers fail to anticipate and accountfor some of the potential benefits,’ The crucialdecisions generally concern pace and priori-ties. How fast should the company move intoexpert systems? Which links in a planned globalcomputer network should be installed first? Ifnothing else, such questions point to the broadgray area separating investments made to con-trol costs (where paybacks can be estimated instraightforward fashion) from those undertakenfor strategic reasons (where uncertainty willbe much greater).

Current priorities for strategic applicationsin the services include on-line access to cus-tomer records for tracking shipments and han-dling inquiries, and database applications formarketing, along with electronic customer or-dering.5 Banks and financing companies have

4See, for example, G.J. Michael and R.A. Millen, “EconomicJustification of Modern Computer-Based Factory AutomationEquipment: A Status Report, ” Annals of Operations Research,vol. 3, 1985, p. 25. Of course, accounting calculations sometimesserve simply to validate decisions made for other reasons.

sFor example, a French chemical manufacturer permits cus-tomer access to the chemical firm’s own database for placingorders and tracking shipments, A chemical company based inthe United States has installed PCs at the water treatment plantsit supplies; the PCs are linked to a host computer for automaticordering and billing. In the future, the host machine will bedirect]}’ linked to process control computers in the water treat-ment plants, These examples come from “Data Processing inMulti-national Corporations,” draft prepared for OTA under con-tract No. 533-6410 by Mackintosh International I.td. ImperialChemical Industries, a British company, likewise plans links with100 of its customers by the middle of 1988 using a commercial\’alue-added network—” ICI Set To Forge Data Exchange LinksWith Its Customers, ” Financial Times, Apr. 23, 1986, p. 10. Notonly purchase orders and invoices, but shipping forms and re-quests for quotes can now be handled over the electronic grid.Electronic pa~ments may follow. The American National Stand-

begun using interactive systems to preparehome mortgage agreements. Hotels provideautomated check-in/check-out services. Secu-rities dealers supply major customers (e.g., pen-sion funds) with terminals that not only displayquotations but enable the customer’s tradersto execute buy and sell orders.

Table 41 lists other examples of strategic ap-plications. These have been divided into twocategories, corresponding to the distinction be-tween micro and macro levels of integrationtreated in more detail in the next section,Briefly, computer integration at the micro levelhas its primary impacts on individual and grouptasks; a work organization perspective becomesappropriate. The macro level of integrationrefers to linkages among departments, divi-sions, and affiliates, as well as to linkagesamong firms; an organizational perspective willgenerally be most useful.

Computer links between firms are not new.Airbus Industrie, the international joint ven-ture, relies on computer networks to coordi-nate engineering design and development, aswell as production, Earlier chapters describedthe SWIFT banking network, along with airtravel reservation systems that interconnect air-lines and travel agents, But in recent years inter-firm computer links have been expanding morerapidly, with ramifications that have yet to beexplored. Certainly these implications seem todiffer from those of telephone or postal com-munications. Netting arrangements among gas-oline suppliers, for example, suggest that thecomputer systems of major oil firms may bemore closely coupled with one another thanwith those of their own wholesalers and dealers,or pipelines and refineries.

Some analysts have argued that, at the mostfundamental level, corporations evolve to min-imize transaction costs—basically, the costs,most of them indirect and less than visible, asso-ciated with moving, manipulating, and other-wise making use of information.6 Continuing

ards Institute has been working on a generic electronic orderform that firms could use regardless of industry–A. Pollack, “Do-ing Business by Computer, ” IVew York Times, July 10, 1986, p. D2.

‘%ee. for example, 0, E. Williamson, ‘‘Transaction Cost Eco-nomics: The Governance of Contractual Relation s,’ ]ournal of

Ch. 8—Moving Toward a High-Sk//l Economy: Computer Applications and Work Organization in the Services • 269

Table 41 .—Examples of Strategic Applications ofData Processing and Communications Systems

Computer integration at the micro level:● An American manufacturing firm coordinates customer

service calls based on a set of software rules that assignpriorities dynamically to the queue of requests.

● A West German bank, which introduced a home bankingsystem in 1983, makes customer account records availa-ble on-line to its sales and marketing representativesthrough 4,800 terminals, (The advantages are greater thanfor an American bank because German banks can hold andtrade in securities for their clients.)

Computer integration at the macro /eve/:●

●

●

Customers can call a fast-food chain’s 800 number forhome delivery. A network of regional centers distributesthe orders among local outlets on the basis of work loadas well as proximity to customers, The chain has equipped10,000 delivery trucks with hand-held computers to helpreduce paperwork, and uses a nationwide database ofdemographic information when planning and promotingnew products,A diversified Japanese MNC provides all of its uppermanagers with on-line access to the corporate databaseon production, pricing, and sales. Experience so far hasshown that some of the managers access the databasehundreds of times per month, others hardly at all,An American-owned book wholesaler uses point-of-sale ter-minals to automate ordering, provide stock control, andspeed response to customer orders.

SOURCE ‘Data Processing (n Multl national Corporations draft prepared for

OTA under contract No 5336410 by Mackintosh International Ltd

technological developments reduce transactioncosts both within firms and among firms. If thenew systems alter or shift transaction costssufficiently, quite dramatic changes in organi-zational form and business practice could re-sult, with major impacts on international com-petitiveness. Note that such outcomes need notdepend on large direct savings in easily meas-ured cost categories. Realignments amongfirms and industries could result from indirector less visible effects of computer-based tech-nologies. For example, shared computer net-works will make communications betweenfirms both easier and more easily hidden. Willinformation sharing lead to collusion and otherforms of anti-competitive behavior?

Already, Chrysler Corp. ’s suppliers can tietheir computers into the automaker’s engineer-ing/manufacturing network, or use terminals.

1,,~ It ,i~](l lj(’ollofnf(’,$,” \’(l], 22, 1 ‘179, p~), 2 33-261; Nl, [{, [Uas\{Jn,

"Transaction Costs and the th[: Thcor} of Nfu]t inat i[)nal [jnter~)ris(>,”

,\’f~tt ‘~’1]~’f)ri{’~ ~)/ ,$/[]//{r),]ti~)~]:/i E~)t(>rprise, A, \l. Rugrna II ( P(1, )II,(jn(l(~n: (;roorn” l { f~lrn, 1082), ~)[~ 24-4:1,

supplied by Chrysler. T GM’s planned world-wide data processing and communications net-work will also link thousands of suppliers, aswell as GM offices and plants. Chrysler and GMshare many of the same suppliers. These sup-pliers may also sell to Ford and Toyota. Indeed,Ford and Toyota may sell components or sub-assemblies to Chrysler and GM. How far willintegration of the various systems go? From atechnical standpoint, the growing consensusaround GM’s Manufacturing Automat ion Pro-tocol suggests that mutual compatibility willbe the eventual outcome. If so, what will theconsequences be, if any, for competition, na-tionally and internationally? At the least, suchpossibilities pose new questions for antitrustpolicy.

Integration: Technical andOrganizational Dimensions

By now, many companies in advanced indust-rial economies have enough experience withcomputers, peripherals, communications links,and software to regard them as standard toolsof business practice, But if familiar, hardware,software, and their applications have grown upindependently of one another, and often remainincompatible. Equipment from different man-ufacturers may not be able to communicate.Users in different parts of a company oftenmake differing technical and organizationalchoices.

Integration at the Macro Level

Worldwide integration requires effectivecommunications among people who may bethousands of miles apart. Firms with manage-rial responsibilities dispersed to many locationsneed systems that can provide effective com-munications both horizontally and vertically(figure 48). Typical applications include:

● finance (reporting by divisions and subsidi-aries; consolidation of accounts; manage-

— ---“ ~ l~r{~~]u~t,~.it~., ~l]a] it~ ~n[~ I]rofitahi]itj Through I{nlerglng

‘1’[)(:hrlologi(;s,”” .~utom~)tjtc EnginccrIrIg, Nlarch 198(;, p, 52, ‘1’he[;h r}~ler ~}stcnl, {:cntert;(] on 28 Iarg[) [)ro(:ess[)rs, in(, III(ics n)[)r”~:than 1, 100 t(:rrninals, ahout }liIlf’ for (.olIli)llter-iii (l(;(l (l{}si~n ,111(1ma nt] fa(, t (1 r i rig.


Figure 48.— lntra-Corporate Information Flows

SOURCE: Mackintosh International Ltd , under contract No 533-6410 to OTA

●

●

ment of both short and long term financ-ing, including international diversificationof risk and cash management);marketing and distribution (coordinationof advertising/marketing strategies; proc-essing of orders; inventory management;pricing; shipping);R&D, product design, and production con-trol (international coordination of R&D

●

projects; computer-aided design and manu-facturing or CAD/CAM databases, includingcustomization of products for particularmarkets; production planning and schedul-ing, including change notices, quality con-trol records, and inventory control for pur-chased inputs as well as final outputs);administrative and clerical (electronic fil-ing; messages and mail; employee travelarrangements).

Many companies began with financial report-ing, but as multinational manufacturers gainedexperience with international data networks,more of their internal traffic has dealt with plan-ning, production control, purchasing, and sales.

During the 1970s and early 1980s, large cor-porations often put together separate networksfor marketing, engineering (perhaps includingproduction), and administration. In manycases, these networks, with beginnings in differ-ent places at different times, remain incompat-ible; IBM currently operates five or more (table42). American Express maintains separate data-bases for its regular and gold cards.

Integration in a technical sense suffers whenone piece of hardware cannot talk to another,and when software developed for one computerwill not run on others. Integration in an organi-zational sense suffers when people in differ-ent parts of the corporation with related jobscontinue to work with different systems. In thepast few years, managers in Japan and Europe,as well as the United States, have begun seek-

Table 42.—IBM’s lntra-Corporate Computer Networks

Network Function

Professional Office Support System Worldwide electronic mail linking 290,000 of IBM’s400,000 employees.

Digital Communications System Radio-based system for 20,000 field service repairengineers in the United States, each of whomhas a hand-held computer for communicatingwith a dispatch center, getting technicalguidance, checking on the availability of parts,and billing customers.

Remote Technical Assistance Links 40,000 engineers in 62 countries, primarilyInformation Network for troubleshooting.

Hands-On Network Environment Sales and marketing system for 25,000 employees.Administrative Access System Ordering and payroll for 35,000 employees.SOURCE D. Kneale, “Sharpening An Edge, ” Wa// Street Journal, Nov 10, 1966, p. 38D

Ch. 8—Moving Toward a High-Skill Economy: Computer Applications and Work organization in the Services • 271—

Photo credit Chase Manhattan Bank

Large American banks maintain officesin dozens of countries.

ing to reduce these incompatibilities and tomore effectively integrate computer applica-tions into ongoing operations. While capitalcosts will be high, many MNCs have decidedthe time has come when the benefits of greatersystem integration (including lower operatingcosts) will outweigh the investment costs.

While desirable in principle, technical inte-gration—so that equipment can communicateeasily, software run on different machines—promises to come slowly at best. Suppliersdifferentiate their equipment for marketing aswell as technical reasons. So long as central-ized batch processing was the rule, few usersmade compatibility a high priority; they sim-ply picked equipment for a given, specific pur-pose, As a result, most users now have substan-tial investments in incompatible hardware andsoftware, inventories that have grown piece-meal over the years. The fast-food chain Mac-Donald’s, for example, operates computersfrom half-a-dozen manufacturers at its head-quarters alone, none of which can communi-cate with the others. e

fI1] [)s Ides ~pf}{ 1 a 1 i ~,[~(1 i~[ I rk \t at it) ns from +[?\e I a I \en (1 () rs, t h[Ic:om [la n ~’ ha\ ~eneral [){] rl)()\(! rna(: h i nc\ bu llt IIJ’ I 13 hl, ‘[’a n(icnl,

The choices of hardware and software a firmmakes as it seeks to integrate its computer sys-tems will depend on the company’s existinglines of business and strategy, but will also helpshape the firm’s future structure. Horizon-tal/vertical integration may be encouraged ordiscouraged. Product lines and manufacturingtechnologies of certain kinds may be favored.Employees with certain skills will become moreor less valuable.

If many of the corporate networks operatedby U.S.-based multinationals remain independ-ent and incompatible, American firms nonethe-less seem to be ahead of their foreign competi-tors at macro level integration.9 They haveinvested more heavily in wide area networks,and use them more extensively and effectively.This lead holds both in high-technology manu-facturing (commercial aircraft, computer hard-ware) and in services (banking, insurance).

U.S.-based service firms have generally movedfurther toward integration in a technical sense(e.g., standardizing on network protocols andsoftware packages) than either European or Jap-anese MNCs. Nonetheless, the decentralizedmanagement typical of American companies,while creating an environment in which inno-vation can flourish, often leads to differing tech-nology choices across departments and divi-sions, as well as across countries. In contrastto American firms, operating divisions in Jap-

and AT&T, plus Wang word processing equipment-j. Rippeteau,“Enter the Intelligent Telephone Line, ” Financial 7’imc. s, Apr. 4,1986, p. 12. Note that many computers cannot commun i(:ate e~erlif built by the same manufacturer.

Ch. 5 included other examples of incompatibilities among s\s-tems, and the growing impetus for integration and commonal-ity}’. Many M NCS view their co mpatihil it~’ p rob] erns as seriousal read~’, and bound to get worse before the~’ get bett~’r. I n gen-eral, a company that imposes compat ibilit~’ standards will facegreater expenses in the short term because d if fereut departmentsmay not be able to choose the low-cost system for their particu-lar needs. Multinationals that hate gro~lrl b}’ a(:quisition faceparticular problems in terms of compatibility}’, but lack of inter-uatio nal standards for hardware and e\en more for softwaremakes integration a long-term goal for everyone. Ch.s. !) ;+n{l I ()illustrate some of the problems from the standpoint of inter na -t iona] agreements on technical standards.

~“Data Prm,cssin,q in hlulti-nationa] Corporations, ” op. cit. Llc-ccntral ized American (:ornpanies experiment more with ne\\ dataprocessing and communicant ions technologies. European multi-nat Ionals tend to bc hehind [1 .S .-based hl NCS in technology a n(ibehind )apanesc hlN(’s at nli(ro-leiel integration of people intothe s}sl[”nl.


anese corporations have less autonomy indeciding on computer applications. Company-wide standards are more common; greater ef-fort goes toward searching for consensus onhow to implement computer-based automation.Japanese companies are slow to adopt com-puter and telecommunications technology, andhave lagged in putting together integrated mul-tinational organizations (as noted elsewhere,their strategy in manufacturing has generallybeen to export from Japan unless or until forcedto invest abroad), But at the micro level of workorganization, their careful and cautious ap-proach has served the Japanese well, at leastin manufacturing industries.

Integration at the Micro Level

U.S.-based MNCs tend to be ahead of theirJapanese counterparts in the use of the mostadvanced hardware and software. But it seemsclear that the Japanese are well in the lead whenit comes to manufacturing systems that effec-tively integrate people and machines—mattersof task design and work organization more thanchoice of equipment.10 This has been a majorsource of competitive success in industries in-cluding automobiles and microelectronics, Oneindicator: the ability of Japanese firms toachieve higher yields and higher quality in theproduction of large-scale integrated circuitswhile using the same equipment as Americanfirms.

The consensus-building mechanisms thathave left some Japanese corporations behindin multinational integration here give them ad-vantages: their ringi decisionmaking processes

‘“See, e.g., International ~’[)xl]~)t:titi~’(?x](?ss in Ele~:tronics, op.c it,, chs. 6 and 8,

lend themselves to conflict resolution and thedevelopment of shared values, necessary attrib-utes of integrated systems at the micro level (forreasons discussed in the next section). In gen-eral, large Japanese companies also invest morein human capital than their American counter-parts—e.g., in corporate training and retrain-ing programs. Thus far, however, there is littleevidence that Japanese service firms gain com-petitive advantages through better integrationof available technologies into ongoing oper-ations.

The question then becomes: If large Japanesemanufacturing companies get a competitiveedge through their ability to design and man-age integrated production systems, will Japa-nese service firms eventually do the same?OTA’s analysis suggests that they will, althoughthe relative immaturity of many Japanese serv-ice industries means that it may take 10 or 15years.

Over the medium to long term, Japanese com-panies should be able to successfully adapt theirconsensus-based organizational traditions toproduction processes characteristic of theknowledge-based services. Japan’s push intosoftware and fifth-generation computer sys-tems, discussed elsewhere, indicates high levelsof resolve in both government and industry. Asthey have done in the past in manufacturing,Japanese companies will probably be able toavoid some of the mistakes made by pioneer-ing U.S. firms. If today their hardware and espe-cially their software remains well behind thatfound in American service firms, there seemslittle doubt that Japanese companies will even-tually make good technical choices, and arriveat production systems well-suited to the char-acteristics of their labor force.

SYSTEMS DESIGN AND PEOPLE’S SKILLS

Data processing and communications sys- view, But the ideal is seldom approached, giventerns in all their variety place new demands on the pace of technological change, the foiblesthe people who use them. Ideally, of course, of designers and their fascination with tech-both hardware and software would be devel- nology for its own sake, the universe of alter-oped with the needs and abilities of users in native system architectures permitted by the

Ch. 8—Moving Toward a High-Skill Economy Computer Applications and Work Organization in the Services . 273

many component and subsystem choices, (See,for example, the section entitled “Technology”in the MetroBank case study, app. 8A, )

Technical capabilities impose one set of con-straints. Some things are possible, some are not;some expensive, some cheap. Managers at adistance from the workplace may impose theirown constraints. Who, then, actually designscomputer-automated jobs? In most cases, tech-nical experts have more control than any othergroup, simply because they are the only oneswho fully grasp what can be done. Given theirmonopoly on technical knowledge, the experts—including those who work for vendors andsuppliers—have a great deal of influence overthe perceptions and expectations of everyoneelse. Sometimes the experts or the managersconsult the people who will have to use the sys-tem, sometimes they don’t. In many cases itseems safe to say that, when it comes to thedesign of work, as opposed to the design of thesystem in a strictly technical sense, no one isin charge.

In such a setting, job requirements evolve,often with a great deal of trial and error. De-spite confusion, uncertainty, and mistakes,large and small, new patterns in the use ofcomputer-based systems have begun to emergein American industry—albeit patterns that arenot yet sharply defined. Flexible product de-velopment and production systems, in the ex-treme verging on customized production, havebecome primary objectives for businesses oper-ating in shifting and unstable environments,domestic and international, in service indus-tries and manufacturing.

Flexibility means different things in differ-ent contexts, but first and foremost impliesrapid response to shifting competitive circum-stance: 11

● tailoring product attributes in response to

I Istudents and practitioners of management, along with critics,

periodical}’ rediscot’er the ~’irtues of flexibility, Simon, \vritin~more than 20 years ago, covers much the same ground as thosein the 1980s who call for flexible organizations and customizedproduction as remedies for the competitive dilemmas of Amer-ican industry. See, for example, H.A. Simon, 7’he Shape of A uto-rnation (New York: Harper & ROLV, 1965).

On flexibility in manufacturing, including applications of ro-botics, see International Competitiveness in Electronics, op. cit,,pp. 233-246.

changing patterns of demand, or to createchanges in demand (the sport shoe exam-ple in ch. 5);thrusts into new geographic or productmarkets (interstate and offshore banking,the home equity lines of credit spurred bychanges in U.S. tax law);new products made possible by new tech-nological opportunities (on-line informa-tion services);shifts in operating level or product mix, asa consequence of business downturns ornew competition (American automakers,hit hard by Japanese competition in theearly 1980s, sought to drive down theirbreak-even points, enabling profitable oper-ations at lower production volumes);changes in government policy (leading,for instance, to new opportunities for serv-ice firms that provide hazardous wastedisposal).

Plainly, technology itself is part of the problem;technical change comes more rapidly than everbefore, and firms in many industries find it hardto keep up. But technology is also part of thesolution. The ability of an organization to re-spond to change depends on its store of tech-nical knowledge, on how well its employees canuse the tools available—whether these are de-veloped internally or purchased in the mar-ketplace.

Sometimes competitive circumstances callfor computer systems that replace peoples’skills, sometimes for applications that enhancepeoples’ skills. In the first case, typified by trans-actional applications emphasizing cost controland illustrated by (most of) the insurance claimsprocessing examples in box Z, the automatedprocess is a relatively mechanical one; the sys-tem does more or less what people once did.(Box C inch. 1 distinguishes between transac-tional and analytical applications of computersystems.) Back-office paper processing in banksprovides another set of examples. In the serv-ices, semi-skilled clerical employees are gen-erally the first to find their jobs deskilled orgiven over to the system.

The second case, enhancement of people’sskills—illustrated by the ways in which Metro-Bank’s CSRs can use their terminals—includes


both analytical and strategic applications. Herethe computer helps people do things they couldnot do before—e.g., interpret signals from aCAT scanner—or helps them do things fasteror more easily. Giving employees in the frontoffice of a bank on-line access to customerrecords creates a new menu of organizationalopportunities. Their work now an intrinsic partof the firm’s marketing strategy, MetroBank’sCSRs had to acquire a new set of skills. Thisentailed much more than learning to use an “in-formation utility. ” The greatest changes wereattitudinal—learning to actively sell the bank’sproducts. The new demands on the CSRs werepart of a much broader set of shifts, as Metro-Bank tried to alter its culture—in part to copewith uncertainties posed by deregulation. Withnew job requirements and new learning comebetter career prospects, should the CSRs wishto take advantage of them (some do not).

Other examples from banking include deci-sion support systems incorporating economicmodels for use in judging lending risks. (Howwill falling oil prices affect a small, independ-ent gasoline distributor’s business?) Again, thecomputer system enhances people’s skills—here people who will probably have profes-sional skills to begin with. To take a differentexample, when computer-aided drafting re-places manual drawing in architecture, engi-neering, and construction firms, productivitygoes up, sometimes by factors of 10 or more.Beyond the direct impacts, designs can bechanged more easily and more quickly; thecomputer can estimate construction costs foralternative designs, prepare bills of materials,estimate heating and cooling loads, prepare per-spective drawings in sun and shadow. Design-ers can explore more options. Clients can pur-sue them in greater depth. The design firm hasgreater flexibility: it can respond more readilyto customer needs and desires, pursue newkinds of business. (Somewhat paradoxically,another result of computer-aided drafting islikely to be greater standardization; the com-puter can store and recall design features froma library, for the operator to put together moreor less mechanically.) But while the automatedsystem opens up new avenues for the designer,

it takes over many of the manual skills of thedraftsperson. Jobs for drafters are deskilled.Companies with these systems commonly hirepeople with vocational-technical schooling butno more than, say, a year’s manual drafting ex-perience. They feel that those with longer ex-perience will be overqualified (and perhapsoverpaid), and unable to adapt as well.

As the drafting example suggests, computerapplications in the services lead to the deskill-ing of some jobs and the upskilling of others.The patterns can be complex and confusing,with many exceptions, but the empirical evi-dence indicates that upskilling will be morecommon where people already have good skillsand good educational credentials. Other jobstend to be deskilled.12 Computers spread know-how through an organization, making it avail-able to many more people, raising the averageskill level at which employees can operate, andhelping to preserve and maintain knowledge;

1zFor a summary, see p. Flynn, “The Impact of TechnologicalChange on Jobs and Workers, ” final report to Department of La-bor, Office of Employment and Training, 1985. Also Technol-ogy and Structural Unemployment: Reemploying DisplacedAdults (Washington, I)C: Office of Technology Assessment, Feb-ruary 1986), pp. 335-354.

Ch. 8—Moving Toward a High-Skill Economy: Computer Applications and Work Organization in the Services ● 275

in the process, they make some people, and theskills these people have, redundant.

The implication is straightforward but daunt-ing: American companies and American edu-cation and training institutions will have to doa better job of preparing people for high-skilljobs, Plainly, many Americans will continueto fill low-skilled, dead-end jobs. But jobs in ter-tiary services like retailing, hotels and restaur-ants, and recreation depend on the size of theNation’s economy, on living standards, and insome sense on international competitiveness.Expenditures on health care, entertainment,and vacation travel go up with levels of afflu-ence; most people would prefer jobs at Blooming-dales to jobs at K-Mart. The more high-skill jobsthe U.S. economy can create, the more workof other kinds there will be. This section, then,touches on the nature of skills themselves, be-fore going on to the ways in which companiesuse the skills of their employees. (Ch. 10 treatsthe implications for human resources policies.)

Well-defined skills characterize some profes-sions, vocations, and occupations, but the no-tion of skill remains fuzzy and ambiguous be-yond rudimentary levels. Box BB touches onsome of the reasons. Reading, writing, andarithmetic as taught in schools are skills. So arereading critically, writing incisively, and think-ing quantitatively (along with learning itself)—but these are much harder to pin down or toteach. And at higher levels, as Polanyi pointsout so aptly, skilled people “know many morethings than they can tell.” 13 Schooling, ofcourse, also conveys the rudiments of work-place discipline—showing up on time, tolerat-ing if not respecting authority. Along with theirother objectives, schooling, apprenticeships,and training programs help people learn to dealwith co-workers, customers, and clients.

The better a firm uses the skills and knowl-edge of its employees, the more competitive itcan expect to be. There is more to this thangrouping tasks as if they were building blocks,just as there is more to computer applications

than distinguishing between replacement andenhancement of people’s skills. But from theperspective of the system and how it functions,two extremes in the design of work can use-fully be distinguished. At one extreme—whenthe computer is used in more-or-less direct fash-ion to automate what people once did—the sys-tem will be rule-based and mechanistic. The“program” is a rigid one, procedures formal-ized, mass production—of insurance claims orModel Ts—the objective. At the other extreme,where product and process characteristics varyand flexibility becomes a buzzword, the or-ganizational program must vary too. This, ofcourse, is one of the things computers are goodfor: flexible rather than fixed automation. Soft-ware can be written to accommodate variation,new programs loaded as needed. On the leftaxis in figure 49 (which is identical to figure6 in chapter 1), the two extremes have beenlabeled adaptive and rigid, suggesting the differ-ence between a system that can adjust dynam-ically to its environment—even if that environ-ment is shifting and unstable—and one that canchange only slowly.

Figure 49 suggests how various industriesand enterprises might be characterized in termsof work organization and computer utilization.Where a firm belongs will depend on patternsof computer use among the occupational groupsin its work force. Table 43 provides a generalframework. The table breaks occupations downinto two major categories: those in which com-puter systems have not (yet) had substantial im-pact on work and skills (the first 3 of the 1 1occupational groups); and those where utiliza-tion of computer systems in the production proc-ess is common and helps define the nature ofthe work (the remaining 8 occupational classes).

Organizations with limited and/or routine useof computers tend to cluster at the left in fig-ure 49, some characterized by rigid forms ofwork organization (fast foods), others by moreadaptive forms (real estate). Most of the jobsin such organizations would fall in occupa-tional categories 1-3 in table 43, with some per-haps in 4-7 (extensive but routine computer use].

A second major cluster of enterprises, at thetop right, consists of those in which computer


Box BB.—Tasks, Jobs,

Even “unskilled” workers must possess a widerange of very real abilities, including some de-gree of problem-solving skill. They must get alongwith co-workers, conform to the discipline of theworkplace. Literacy may or may not be neces-sary, but communication certainly is. Skilledworkers rely on broader and deeper stores of tacitknow-how (anticipating problems, troubleshoot-ing, generalizing from a limited number of cases),along with well-codified knowledge (how to useDO statements in Fortran programs). Most pro-fessionals depend heavily on book learning, butneed tacit knowledge and good judgment as well—skills that go far beyond knowledge acquisitionand reasoning.

Some tasks are simple and fixed, repeated inthe same sequence every few minutes or few se-conds. This is the case for much unskilled andsemi-skilled work: data entry, collecting bridgetolls from motorists. Skill levels go up as judg-ment and experience come into play; supermar-ket checkers must be better at making changethan toll collectors. Sales clerks in departmentstores need to assist customers as well as taketheir money. Social skills come into play inalmost all jobs. Retail chains train their salesclerks in how to behave toward customers, Po-lice officers need good sense when it comes topeople and their behavior. In the professions,education and training shape social skills andattitudes (so that physicians, for example, behavequite differently among themselves than they doin front of nurses or patients).

For skilled work, the universe of possible tasksand procedures (hence the tool kit of skills, thebase of knowledge and experience) will be large.The surgeon or cabinetmaker selects from theseas needed to do his or her job. Some of the skillsare manual, some mental. Lawyers may needgood instincts, the ability to think on their feet(like the police officer), coupled with the concep-tual skills and fortitude to deal with complexcases that may last for years. A software engi-neer designing a large program may face thou-sands of choices in arranging instructions,branches, and subroutines; arriving at the over-all shape of the program takes a different set ofskills than writing reasonably error-free code ona line-by-line basis.

As such examples suggest, only relatively rou-tine work can be viewed in terms of procedures

and Skills

put together from sequences of well-definedtasks. The greater the need for conceptual andjudgmental skills, the less precision the notionof work organization conveys.

By the same token, outputs are harder to evalu-ate when work depends on higher-order skillssuch as planning. Standards of quality becomesubject to debate and disagreement; as discussedin chapter 2, consumers may have little basis forevaluating the services of physicians or lawyers,even after delivery of the service. Evaluation mayitself demand judgment and skill (thus creatingjobs for people such as music critics). Distinc-tions between minor league and major leaguebaseball pitchers, chess masters and grandmasters, artists (including computer program-mers and architects) whose work will last or dis-appear, may escape observers who are not them-selves highly skilled. Instinct, feel, judgment,intuition, inspiration—this is the vocabulary ofsuch distinctions. Measures of productivity maybe equally uncertain. Baumol has often noted thatstring quartets performing for alive audience areno more productive today than a hundred yearsago; one might add that people’s ability to evalu-ate and appreciate such performances has prob-ably not improved either.1

Measures of skill, then, can rarely be very pre-cise beyond some point of relatively ordinarycompetence. For such reasons, it is too much toexpect expert systems to be able to replicate theprocedures of people who truly are experts,though it is not too much to expect computer pro-grams to be ordinarily competent.2 When learn-ing, people follow rules and instructions for play-ing a violin or writing software. Those who havebecome acknowledged experts may sometimesfollow clearly visible rules and procedures, some-times not. Sometimes they break the rules.Higher-order skills involving problem-solving

%M, for example, W.J, Baumol, “Productivity Policy and the Serv-ice Sector,” Managing the Service Economy: Prospects andl%oblems,R,P. Inman (cd.) (New York: Cambridge University Press, 1985), p,301. Beyond the live audience case, broadcasting and recording tech-nologies have of course led to huge productivity advances,% H. Dreyfua and S. Dreyfus, “Why Computers May Never Think

Like People,” Technology Review, January 1986, p. 43; alao H.M. Col-lins, R.H. Green, and R.C, Draper, “Where’s The Expertise?: ExpertSystems As a Medium of Knowledge Transfer,” Expert Systems 85,M. Merry (cd.) (Cambridge, UK: Cambridge University Press, 1985),p. 323.

Ch. 8—Moving Toward a High-Ski/l Economy: Computer Applications and Work organization in the Services . 277

normally come only with time. Extended train-ing, with a good deal of supervised practice, gointo making a competent physician or air trafficcontroller; expertise follows with accumulatedexperience.

Finally, it is seldom that isolated tasks set thebounds for work. Many jobs require a good dealof contextual knowledge—an understanding ofhow the organization and its processes function(table 43), An employee with good contextualknowledge should be able to diagnose problems,

Figure 49.— Characteristics

Adaptive,Flexible

Organizationof

work

Rigid,Pre-programmed

dynacomparadvant

help customers, and otherwise get things donebecause he or she knows where to go and whoto talk to; relationships with other people mayin fact be the primary defining features of com-puter-assisted jobs. The associated skills may alsobe the most difficult to learn: the changing inter-actions between bank employees and their custom-ers discussed in appendix 8A—and the changingrelationships between supervisors and subord-inates—illustrate some of the complexities of de-veloping new social and managerial skills.

of Firms and Industries

Enhancementof human

skills

Tertiary I Old New / Knowledge-

Computerutilization

replacementof human

skills

(traditional)/

manufacturing manufacturing/

basedservices services

SOURCE Office of Technology Assessment 1987


Table 43.—Patterns of Computer Use by Occupation

Occupations with limited and routine use of computersystems

1. Shopfloor work in manufacturing where computer-based automation is uncommon.

2. Service industry jobs with little or no use of computers(hair stylists, entertainment).

3. Managerial and professional work where people use thesystem at their own discretion, or for routine functionsonly (e. g., word processing).

Occupations with extensive computer utilization

Routine:4. Input-output (data entry, materials handling, loading and

unloading of automated production equipment). In gen-eral, input-output jobs require little or no interactionwith customers, in contrast with data gatekeeper jobs,below.

5. Data gatekeeper—jobs at the boundary between thefirm and its customers or suppliers, but requiring lit-tle contextual knowledge of the product or productionprocess (e.g., directory assistance operator, order takerfor classified advertising).

6. Machine tender (routine monitoring of computer-auto-mated equipment).

7, Troubleshooter (component or equipment level, includ-ing routine maintenance and repair, as well as qualitycontrol).

Non-routine:

A. Technical and paraprofessional8. Customer/client/intrafirm service representative—pro-

vision of services, outside or inside the firm, requiringgreater contextual knowledge than data gatekeeperjobs (e.g., loan officer, travel agent, insurance adjus-tor, administrative assistant). In jobs with high contex-tual knowledge requirements, people must understandhow the organization works and be able to use thatknowledge to get things done.

9. Diagnosis, evaluation, repair, operation, system main-tenance (database librarian, paralegal, medical labora-tory technician, currency trader).

B. Design, system control, professional10. Process control, system troubleshooting, crisis inter-

vention (air traffic controller, powerplant operator,financial manager, numerical control machine toolprogrammer).

11. Product, process, and system design (urban planner,architect, chemical engineer, research scientist).

SOURCE Adapted from ‘International Competition in the Service Industnes Impacts of Technological Change and International Trade on U S Employment, ” prepared for OTA by E Appelbaum, P S Albln, R Koppel,and F Hormozl under contract No 533-5560

systems enhance the skills of employees inadaptive settings. The greater the fraction ofa firm’s jobs that fall in categories 4-11 of table43, the farther to the right the firm would beplaced in figure 49. A preponderance of jobsin categories 4-7 implies using the computeras a replacement for human skills, hence a po-sition in the lower right portion of the chart.More jobs in categories 8-11 would move thefirm upward, closer to the pole marked enhance-ment of human skills.

Firms with rigid forms of work organizationtypically employ many people in routine occu-pations, complemented by a small superstruc-ture of technicians and professionals. Adapt-ive organizations, on the other hand, generallyneed people with both broader and deeperskills, and employ relatively large proportionsof workers in technical and paraprofessionalcategories. Rigid work organization correlateswith deskilling (except for the technicians andprofessionals who design the system and keepit running), adaptive with upskilling.

As figure 49 and table 43 suggest, among theoccupations most susceptible to technologicaldisplacement are routine input-output and datagatekeeper jobs with low contextual knowledgerequirements. For instance, self-service, aidedby computer systems, has meant declining jobopportunities for gasoline station attendantsand bank tellers. Other service jobs requiringlittle contextual knowledge can be exported tolow-wage offshore locations (ch. 7). In contrast,high contextual knowledge implies judgmentalskills; such jobs are harder to automate, al-though offering many opportunities for com-puter enhancement and assistance. Also notethat tenure in jobs with low contextual knowl-edge requirements does little to prepare work-ers for upward moves within the firm.

INTEGRATED PRODUCTION SYSTEMS: THE NEW MODEL

Figure 49 suggests the outlines of a new services and goods, emphasizing strategic ap-model for computer-assisted production, one placations of computers and patterns of use thattypified by U.S.-based firms in the upper right enhance the skills and abilities of at least someof the picture. These new-model firms design, of their employees. Many have organized them-develop, produce, and market knowledge-based selves as integrated production systems, with


integration implying substantial use of com-puters and telecommunications to more closelycouple product design, production, and mar-keting and to tie together dispersed geographiclocations. In search of flexibility and entrepre-neurial behavior, these companies tend to stressemployee training and skill development, alongwith controlled risk taking, delegation of au-thority, and decentralization. They often ap-proach their markets on a global basis, evenwhen small (e. g., PC software firms).

Some manufacturing firms, as well as knowl-edge-based service suppliers, fit the new model,In both the services and manufacturing, thenew model differs from the old in the socialsystem as well as the technical system, withmanagers recognizing that it is the interrela-tionships of computers and people, more thanthe technical system in isolation, or people’sskills and abilities in isolation, that determinesthe firm’s ability to respond to changing cir-cumstance, and, over the longer term, to meetinternational competition.14 New social systemsgenerally attempt to increase employee iden-tification with the organization, and commit-ment to its goals. Sometimes this means realif localized power for employees at lower levels—e. g., a voice in hiring new members of workgroups (and perhaps in firing), peer approvalfor pay raises, Sometimes it means no morethan consultation on changes in work organiza-tion. In other cases, it has become little morethan a sophisticated form of paternalism.

A greater fraction of the employees in new-model firms will need specialized knowledgeand training, sometimes job- or occupation-specific, sometimes firm- or industry-specific.Computer graphics, for example, has narrowed

1 qrrh is ~or~espon~s to the classical statement of objectives ofthe soc iotechnical systems approach to job and organizationaldesign-the search for a joint optimum between the technicals}’~tem (including both process and product technologies) andthe social system (that is, the people in the organization, withtheir skill endowments and foibles, preferences for cooperationor (,[)nfli( t. informal groupings). See, for example, E. I,. Tri\t,et a]., Organizational Choice: Capabilities of L’rou[).s at the CoalFace 1‘rider L’hanging Technologies (London: Talistock, 1963];E. I.. ‘[’rlst, ‘ ‘The Evolution of Soc iotechnical Systems as a Con-(,e~)t ua 1 F’ra mewo rk and as a n Action Research Program, Pcr-specti~es on organization lle.~ign and Beha ~ior, A. H. L’a n d et’en an(l tl’, F, Jo}(.[’ (e(ls. ) [NeY\ }’ork: 1}’ile}, 1981), p. 19.

some of the differences in skills among peoplewho do cartoon animation, design theater andmovie sets, draw maps, or prepare graphic lay-outs for magazines and books, financial dis-plays, scientific illustrations. Such jobs arefound in different industries, and traditionallyhave been viewed as requiring different skills.With computer automation, they are converg-ing. Likewise, data gatekeeper jobs call on sim-ilar kinds of skills across industries—commu-nicating over the telephone, customer service,using data processing systems to access, store,and retrieve information. Note that overlap injob skills among seemingly different occupa-tions and industries eases lateral mobility, onereason for growth in temporary employment(ch. 7). In general, employees of new-modelfirms will be expected to bring broader (andperhaps deeper) skills to the workplace; thecompany may use job rotation, work sharing,and small-group production to help develop amulti-skilled labor force. Contextual knowledgewill be more important for more people, whowill need to know how their work fits into thechain of production, so that they can trouble-shoot problems, help customers, make good de-cisions.

Somewhat more concretely, most companieswith organizational structures evolving towardthe upper right portion of figure 49 share char-acteristics from the following list:15

● Firms define jobs somewhat more broadlythan in traditional organizations. Some-times broader skills and responsibilities fol-low more or less directly because computerautomation permits each person to do more.Sometimes groups take over responsibil-ity for a number of tasks, with individualslearning several jobs and rotating amongthem. By working in groups, people learnfrom one another, more easily share skillsand information. When people know eachothers’ jobs, or can simply do more, theorganization becomes more flexible.

15A n earlier \,ersion, adapted for manufacture ing, appeared i nTechnolog~’ and Structural Unemplo~’ment: Reemploj’ing Dis-placed Adults, op. cit., pp. 356-357,


●

●

●

New-model firms, where an employee’sactions may have impacts that reachdeeper and spread more broadly than ina traditional organization, call for greatercontextual knowledge. With many centersof responsibility, and people at relativelylow levels making decisions and taking ac-

tion, employees need to understand h o wtheir work affects the rest of the organiza-tion. When the same person sells insuranceand prepares the policy documents at an

on-line terminal, that person must knowmore and take more responsibility.

When people gain more responsibilityand control over their work, they may getmore satisfaction from their jobs (some do,some don’t).As a result of broader job definitions andthe need for contextual knowledge, corpo-rations may find they need to give their em-ployees more and better training, and mayseek deeper as well as broader skills (evenamong employees normally classed as un-skilled or semi-skilled).

Firms that have moved to product-cen-tered rather than functional organizations(a distinction discussed at greater lengthbelow) usually provide training coveringthe product line; a few companies h a v egone beyond, say, marketing information,beginning to discuss profitability or long-range planning with their employees—topics ordinarily reserved for managers.Many training programs are intended to

expose employees to corporate goals andenhance motivation, sense of belonging,and commitment to those goals. Such ob-jectives often merge into the developmentof contextual knowledge; for instance, a

company may devote considerable effortto showing employees how their jobs con-tribute to the firm’s end products. Peoplemay be encouraged to view themselves asimmediate participants in the productionof final services, rather than simply doinga job somewhere along the chain of pro-duction, (“I work for MetroBank” ratherthan “I’m a bank teller.”)Employees at lower levels maybe granteda say in decisions on equipment and pro-

cedures (e. g., word processing or spread-sheet software), as well as day-to-day oper-ations.

Typically, participation takes the formof meetings between employee represent-atives and the company’s technical andmanagerial staff. Planners may seek todraw on the experience of the current workforce.

Consultative mechanisms, regardless ofform, seldom give low-level employees anyreal control over major decisions—thosethat shape the system. Only in cases of mas-sive opposition to proposed changes, o r

where employees are represented by stronglabor unions with independent sources o f

technical expertise, are they likely to b eable to influence the fundamental choicesthat shape the organization of work and

the application of technology.Managers may give groups of workerssome or all of the authority formerly vestedin first-line supervisors, perhaps includinglimited control over pace, task design, andwork methods, along with responsibilityfor quality and for coordination with otherdepartments.

When groups take over responsibility formonitoring absenteeism, for allocatingwork, and for quality control, the supervi-sor’s role may become primarily that offacilitator and communications channelwith higher management, As illustrated inappendix 8A, supervisors commonly findthemselves spending more time on man-agement, less on administration. Often, theratio of supervisors to production employ-ees declines; some companies have elimi-nated first-line supervisors entirely.

Giving supervisory control to workgroups can heighten job stress. Among thecauses are intra- and inter-group competi-tion—forces that managers look to forgreater productivity. Moreover, work groupsoften have some control over membership—perhaps the power to veto prospectivenew employees. People who do not fit inmay find themselves not only uncomfort-able, but out of a job, Work groups carrypotential for inequities and abuse that few


●

●

●

American companies have as yet ac-knowledged.In selecting new employees for some kindsof jobs, companies may weigh motivationaland attitudinal factors more heavily thanpast experience (or, in some cases, educa-tion). Social skills may get new emphasis.

Some American firms have adopted mul-tiple levels of screening, with aptitude andperhaps psychological tests followed by in-terviewing. The interviews might involveprospective peers as well as the personneldepartment and supervisors.Pay scales may reflect the skills an em-ployee has acquired (e.g., the jobs he or shehas mastered—pay for skills) and/or per-formance (payment for results). In additionto meeting objective standards such aswritten tests, an employee seeking a pay-for-skills increment may have to be passedby other members of the group, as well asby supervisors.In decentralizing, some companies havereplaced functional with product-centeredorganizations. The common objective is tochannel work smoothly and directly frominput to output of the system—creating afaster, more flexible (if not necessarilymore efficient) production process.

The benefits of “channelization” can bequite real, even if hard to quantify. As dis-cussed below, banks have replaced central-ized data processing divisions with smallerdepartments specialized by type of cus-tomer (retail stores) or transaction (cur-rency exchange). When a single depart-ment provides most or all of the servicesa given client needs, the bank’s employeescan respond more quickly, the system be-comes more nearly transparent to the cli-ent (and to the bank’s own employees).

Product organizations can also contrib-ute to employee motivation and commit-ment; people may identify more readilywith a department that supplies a completeproduct rather than a piece of one.

people. This may be a matter of circumventinga lack of literacy or numeracy in the labor pool,designing equipment so that it can be used moreeasily and more productively (user-friendlycomputer systems), or finding better ways tocall on the capabilities of those who are over-educated for the jobs they find themselves in,

Although no census exists, several thousandU.S. companies appear to share a substantialnumber of attributes from the list above, withthe number in manufacturing probably exceed-ing that in services.16 A much larger numberof American firms have taken smaller steps,such as the introduction of quality circles. Inthe services, new model organizations are farmore likely to be found in, say, banking, thanin construction,

As more organizations consider alternativessuch as those outlined above, the questions forU.S. companies and their employees include:Compared with foreign competitors, will Amer-ican service firms react slowly and uncertainly?Or will they take the initiative? How will deci-sions taken by corporations, collectively, affectcompetitiveness? How will skills be affected?Will people be integrated into the system or outof it? Will job ladders and prospects for upwardmobility be truncated?

The answers depend mostly on whetherAmerican managers in large numbers searchout new ways of maximizing the contributionsof individual employees to competitive per-formance—a search that some U.S. firms haveembarked on, along with a few European com-panies and many Japanese organizations (thelatter particularly in manufacturing). It seemsclear that widespread adoption of at least somefeatures of the new model can help improvethe competitiveness of American industries,and thus aid in preserving job opportunitiesover the longer term. Nonetheless, in the shortrun, work reorganization in conjunction withcomputer automation typically causes some

While the primary thrust in most new-modelfirms has been to select and fit people to exist-ing or redesigned jobs, jobs can also be fit to

l~see R, E. Wa]ton, “From Control to Commitment in the \Vork-place, ” Harvard Business Re}riek$’, hlarch-April 1985, p. 76. Also,Product j~’jty Through Work Inno\’ation.~: A tl’ork in AmericaInstitute Polic~ Stud~’ (New York: Pergamon, 1982], p. 3.5.

63-527 0 - 87 _ 10


kinds of job opportunities to disappear, peo-ple to be displaced.

While a number of the steps outlined abovemay contribute to making some kinds of workmore satisfying, new forms of production sys-tems also bring new dangers, born of the in-herent conflicts of interest between firms andtheir employees. Training programs can endup resembling indoctrination. And what, forexample, of people who find themselves at oddswith a work group? The more authority man-agers cede to such groups, the greater the po-tential for abuse by the group. (Which is notto imply that managers never abuse theirpower.) Questions of equity would seem to ex-ist already, first and foremost where groups canveto new employees. The modern, bureaucraticcorporation, after all, emerged in part from ef-forts to counter the favoritism, nepotism, andcorruption that afflicted earlier forms of orga-

nization. In 19th century U.S. factories, Irishgang bosses hired other Irishmen. Groups ofCornish miners working on contract in the Mid-west, if not members of the same family, oftencame from the same village. Work groups todayare hardly likely to take such forms, but possi-bilities for coercion and abuse can plainly arise.

Furthermore, practices such as screeninghundreds of applicants before picking a fewdozen will strike many Americans as less thanfair; a corporation, after all, is not an athleticteam. Labor unions, where they exist—and theyare relatively rare in U.S. service industries (ch.7)—can help safeguard the rights of individualsand protect against discrimination in hiring andfiring, But some American managers see or-ganizational change as a way to keep out laborunions, and not a few companies view unionsas incompatible with new forms of work orga-nization.

IMPACTS ON COMPETITIVE ABILITY

As noted earlier, American managers haveoften been criticized for placing higher priori-ties on immediate cost savings and short-termprofits than on investments leading to indirectand/or strategic payoffs visible only over thelonger run, As this section will show, many ofthe most significant impacts of new-model pro-duction systems come about indirectly. For in-stance, greater job satisfaction can help reduceemployee turnover, hence costs for trainingnew employees (including lost production dur-ing on-the-job learning). Thus the speed withwhich American companies move toward new-model organizational forms may provide onetest of the proposition that incentives in the U.S.economy skew managerial decisions toward theshort term.

Citibank’s shift in its back offices from a func-tional to a channelized or product organizationillustrates some of the indirect impacts, herethrough improved customer service.17 During

17R. W, Wa]ter5, “The Citibank Project: Improving ProductivityThrough Work Redesign,” The Innovative Organization: Produc-tivity Programs in Action, R. Zager and M,P. Resow (eds.) (NewYork: Pergamon, 1982), p. 109.

the 1960s and into the 1970s, Citibank, like mostfinancial services firms, fed all its transactionsthrough a large centralized data processing divi-sion—the functional approach common in anera of batch processing on expensive mainframecomputers. In Citi’s DP division, all transac-tions (e.g., checks) passed in sequence throughhalf a dozen departments organized by func-tion (e.g., coding and keypunching—the systemroughly paralleled that described in box Z forbatch processing of insurance claims). Differ-ent transactions—checks, letters of credit, for-eign exchange—took different routes throughthe DP division, but all were handled in basi-cally the same way, There were no distinctionsbetween, say, corporate and individual ac-counts. Each time the paperwork for a trans-action crossed the boundaries between depart-ments, it moved from the province of onemanager to another, With responsibility frag-mented, errors were easy to make and hard totrace.

The company redesigned and decentralizedits transaction processing system so that a sin-gle manager would have end-to-end responsi-


Photo credit Burroughs Corp

Automated reader/sorter for processing checks.

bility for a single type of transaction (product).Different departments now handle letters ofcredit, depending on the customer (govern-ment, correspondent banks, Citicorp branches)as well as the geographic market. Each trans-action follows a well-defined path or channelfrom the time it enters the Citibank system un-til it leaves, with management responsibilityclearly defined at each point.

J u s t a s i n t h e i n s u r a n c e i n d u s t r y , c h e a p c o m -

p u t i n g p o w e r , s p e c i a l i z e d w o r k s t a t i o n s , a n d

f r i e n d l i e r s o f t w a r e m a k e p r o d u c t f o r m s o f o r g a -

n i z a t i o n p r a c t i c a l i n b a n k i n g . O f c o u r s e , a s s i g n -

i n g d i f f e r e n t d e p a r t m e n t s f u l l r e s p o n s i b i l i t y f o r

t h e n e e d s o f d i f f e r e n t g r o u p s o f c u s t o m e r s i n -

e v i t a b l y m e a n s d u p l i c a t i o n o f e q u i p m e n t . C a p -

ital costs may rise. But today, in many banks,checks stay within one department, lines ofresponsibility are shorter, and, in principle,service is faster and more responsive. A cus-tomer with a question should get a quick an-swer. Even if the company’s cost structure doesnot show improvement, better service shouldbe good for business and for competitiveness.

Given such examples, it should be no surprisethat little concrete data exists for evaluatingwork reorganization and computer integrationin new-model organizations. Certainly, most

companies, when they introduce new forms ofproduction systems, do so to improve effi-ciency, defined broadly. But they seldom re-veal quantitative information on the success ofthese efforts. Indeed, it can be hard for the com-pany itself to tell whether new forms of workorganization have been successful, if only be-cause conventional accounting measures mayfail to capture the full range of benefits. Afterall, what is the dollar value of better customerservice or greater flexibility? How can projec-tions into the future capture strategic ad-vantages?

Some generalizations, nonetheless, hold quitebroadly. New-model organizations normallyhave somewhat higher fixed costs comparedwith old-model systems. Even in the absenceof heavy investments in computers, peripherals,and software, capital costs typically go up forthe simple reason that many job redesigns en-tail redundant equipment for parallel productdepartments or work groups. New buildingswill cost more to the extent that they providemore space or need coaxial cabling and lightguides for computer networks and telecommu-nications links. As table 44 indicates, direct la-bor costs may go up or down, with substantialindirect cost savings possible. Training costsper employee increase when jobs become broaderor more complex, although net expenditureson training may drop if turnover declines. Re-ducing or eliminating first-line supervisors cutslabor costs directly. Shorter, more direct inter-personal communication channels can alsosave money.

But costs give only part of the picture. WhenMetroBank (app. 8A) reorganized, it acceptedrelatively high costs as part of its new strategy.purchasers who get more responsive servicemay order more; new customers may be at-tracted by word of mouth. The point is a gen-eral one: work reorganization can help compa-nies initiate and pursue new business strategies.There is every reason to believe that, over thelonger term, new-model firms in U.S. serviceindustries (and in manufacturing) will be ableto gain advantages in both domestic and inter-national competition.


Table 44.—Typical Changes in Cost Structure for New-Model Organizations

Direction of cost shift

Capital costs:Buildings and equipment Up (smart buildings, more computers)

Transient costs associated with work reorganization:Start up

Lost time for meetings andconsultations with employees duringthe design phase

Consulting and contract services

Direct operating costs:Direct labor content per unit of outputPay scales

Quality control, inspection and error-correction costs

Indirect operating costs:Downtime (unplanned idle time)

Lost production during meetingsTraining costs per workerHeating, lighting, other utilities and

servicesMaintenance costs

Indirect labor and supervisory costs perunit of output

May go up or down. Broadening of jobsand skills tends to increase startupcosts

up

Up, if used

May go up or downUp or down. Pay-for-skills, bonus, and

gain-sharing tend to raise wage levels,but many of the companies in theUnited States adopting such plans willprobably remain non-union.

Down

Down, especially where more flexiblesystems replace old-model organizationalforms such as assembly lines, with theirpotential for shutdowns caused byminor problems.

upupUp or down. Employee suggestions may

lead to savings.Generally up. (More equipment means

more maintenance, although computerand telecommunications equipment isgrowing steadily more reliable.)

Down

Cost categories with both direct and indirect components; intangible costs:Cost penalties and avoidable mistakes Down

resulting from insufficient employeeinvolvement in the initial design ofthe production system

Costs associated with employee Downabsenteeism and turnover (includingrecruitment expenses, training,overtime, wages for relief and utilityworkers)

Added costs associated with operating at Downpart-capacity

Communication costs within the Up or downorganization

Wastage, including theft DownGrievances DownCosts associated with minor changes in Down, unless new equipment needed

product designCustomer dissatisfaction DownOther costs attributable to employee Down

dissatisfaction or low moraleSOURCE Office of Technology Assessment, 1987

Ch. 8—Moving Toward a High-Skill Economy Computer

CONCLUDING

Applications of technology matter for com-petitiveness in services: they exert direct andindirect impacts on costs; lead to new strate-gic opportunities; help ensure consistent andhigh-quality output; keep customers satisfied;enhance the firm’s reputation. Most of the earlyuses of computer and communications systemsfocused on cost cutting. Today, strategic con-siderations—seeking to establish new products,push into new markets, provide better serviceto customers—have overtaken cost control asa managerial priority. Reorganized and moreflexible production systems have become in-trinsic elements in the strategies of a growingnumber of American service firms. Japanesecompanies have turned effective work orga-nization into a potent source of competitiveadvantage in manufacturing; they could even-tually do so in the services, American servicefirms cannot afford to fall behind.

Given the unstable operating environmentsthat now characterize so many internationalindustries, greater flexibility-in part througha labor force with broader and deeper skills (andin part from greater use of temporary and part-time employees)—can enhance competitive-ness. So can technology that helps people usetheir skills effectively, New possibilities for theapplications of technology often appear fasterthan people can apply them. Managers havebeen confused, unsure of how to invest theirfirm’s resources, Lower level employees havebeen confused, unsure whether or not to wel-come the terminal that appears on their deskone morning. Economists have looked at thestatistics and concluded that much of themoney has been wasted, that new technologieshave been deployed unwisely because produc-tivity growth in the services remains low; theyare confused because their measures fail to cap-ture changes in output characteristics.

This avalanche of technology opens newavenues for designing production systems thatmatch the needs of the people using them—and the needs of customers, By adopting sys-tem design criteria that judge technology interms of its contributions to the system as a

Applications and Work Organization in the Services • 285

REMARKS

whole, companies can—in principle—achievecombinations of flexibility and efficiency un-heard of when computers were expensive,Good solutions to the problems of computer uti-lization in new-model production systems cancut costs and improve productivity while alsoyielding strategic advantages.

New-model production systems will be simul-taneousl y more integrated and more flexible,What does this imply? On a macro level, greateruse of computer networks to link dispersedoperating units, and greater commonality i nequipment and in databases—the goal being totie product development and marketing moreclosely to the production process, On a microlevel, it implies decentralized and semi-autono-mous forms of work organization—product de-partments, end-to-end production channels,employees with broader and deeper skills, jobrotation.

In theory—practical implementation is anothermatter, as the appendix to this chapter shows—a more integrated system should (perhaps some-what paradoxically) be able to adapt more read-ily to new product designs, to shifts in mix orvolume, even to changes in government pol-icies, To do so, the system must effectively inte-grate people and machines, which, in turn,means that employees at all levels will need newskills—reasoning, problem-solving, acting au-tonomously.

Micro level integration becomes easier withsmaller, cheaper, more friendly computer sys-tems. For instance, fourth-generation languagesmean that end-users with relatively little train-ing can create their own applications programs,Indeed, fourth-generation software, makingpossible fluid and evolving system designs, maypermit firms to move toward organizationalstructures, as well as particular applications,that can evolve over time to meet the needs anddesires of individual employees,

I n g e n e r a l , e f f e c t i v e i n t e g r a t i o n i m p l i e s g i v -

i n g e m p l o y e e s a t a l l l e v e l s g r e a t e r r e s p o n s i b i l -

i t y , T h i s m a y m e a n g i v i n g t h e m a s t a k e i n t h e

s u c c e s s o f t h e e n t e r p r i s e — a n d i n t h e s e c u r i t y


of their own jobs—through gain-sharing, bonus,or profit-sharing plans. Even among the mosttradition-bound American managers, manyhave begun to acknowledge the need for greateremployee involvement as companies move intohigh technology and face new competitivepressures,

Given better computer and communicationssystems, and a better educated, better trained,and more responsible work force, companiesshould be able to maintain effective internalcoordination and control, even as managementstyles loosen, Why? Because managers seektight control in part as a hedge against partialor faulty information. With an integrated pro-duction system, enlightened corporations mayloosen the reins, upper managers lose some oftheir fear that the system will go out of control.

Meanwhile, less enlightened companies maymove in the opposite direction, towards morerigid forms of work organization, Some willseek to use more and better information totighten supervision, losing the benefits of flex-ibility. Of course, productivity in some indus-tries will continue to depend on close controlover task assignments, and on rigid forms ofwork organization: there is little need for flexi-bility (outside well-defined bounds) in most fast-food outlets, or in a telephone company, And,although managers may seek to reduce costsassociated with employee absenteeism, turn-over, even sabotage, they do not develop newproduction systems just to keep people happy,Particularly in the tertiary services, tensionsbetween productivity and quality of workinglife may change little if at all; managers willpush for efficiency, employees for less stress,a more tolerable environment. Productivity willcontinue to come at the expense of job satis-faction and quality of working life, if only be-cause some kinds of production—in the serv-ices and in manufacturing—will continue todemand routine and repetitive work; flexibil-ity is not always needed,

More broadly, however, continued develop-ment of new-model organizations seems likelyto slowly extend the bounds of employee con-trol. The reasons are largely technological, in-

herent in the design and operation of complexsystems. Distributed computing will be a pri-mary tool for automation and integration.Highly automated systems may need fewer em-ployees, but these employees will bear heavierresponsibilities. Automated systems cannot beidiot-proofed; indeed they tend to be more sen-sitive, less robust than labor-intensive systems.18

One person’s mistake can shut down a highlyintegrated system, where in a more labor-inten-sive organization the consequences would re-main localized. If integrated systems are tofunction effectively, employees must have con-siderable freedom of action, and the knowledgeand skills to intervene swiftly and appropri-ately. In consequence, some of the distinctionsbetween managers and other employees willnarrow. As ordinary employees take on greaterresponsibilities, companies will find it in theirinterests to treat and train them more like su-pervisors and professionals. A high-skill econ-omy will depend on a labor force that can ac-cept authority, use good judgment.

The technology of integrated production sys-tems, then, will break down some of the bar-riers in hierarchical organizations. In manycases, the barriers will come down fastest andfarthest within management. Natural enemies—marketing departments and back offices,managers on the same level in the hierarchy—will have to work together, much as produc-tion and quality control must be integrated innew-model manufacturing plants. Particularlyat middle levels, managers will lose power (andsome will lose their jobs). None of this will hap-pen without a struggle. American industry car-ries a heavy burden: hundreds of thousands ofmanagers who learned the ways of the 1950sand 1960s, many if not most of whom will neverfeel at home in a new-model system.

la’I’he reasons begin but hard]y end with the elusiveness of soft-ware errors, In general, control models themselves can onl~ rep-resent a portion of the system, so that people must bridge theresulting islands of automation. Furthermore, control modelstend to break down because of contingencies that were o\rer-looked or simply cannot be incorporated into the model. By def-inition, when the model breaks down the system is out of con-

trol and people must take over, Technical constraints inherentin control models mean that development of large-scale computer-integrated systems will continue to be slow and painful, withprogress depending on trial-and-error arid incremental im-provement.


In the end, it will be more difficult to tracepower and authority to particular people ex-cept at the upper executive levels. The impli-cation: evolving organizations will need newmechanisms for resolving conflicts amongmanagers with different goals and priorities(one of the things Japanese companies are goodat), as well as for resolving conflicts betweenlabor and management. Where these mecha-nisms do not develop, competitiveness will suf-fer. Likewise, where companies rely too heav-ily on computers, particularly in routineapplications, they may find their work forceslowly losing the core of experience-basedlearning that will always be needed, A furtherimplication: group responsibility will have toreplace individual responsibility, managers willhave to begin thinking of themselves as mem-bers of multi-skilled work teams.

What does this mean for Americans in thelabor force, or ready to enter it? In some cases,it will mean greater stratification within service-oriented companies. Some highly skilled (andhighly motivated) employees will have relativelysecure jobs, with good pay and prospects forrapid advancement, Others will have full-timejobs but limited upward mobility, A third groupmay find themselves limited to part-time andtemporary work, serving as buffers against un-certainty and market fluctuations. Credentialswill become more important for securing thekinds of entry-level positions that lead to on-the-job learning and skill development, Morecareer ladders will begin with post-high schooleducation—2-year and community colleges, vo-cation-technical schools, bachelor’s level pro-grams. High-school graduates—and those whohave not graduated–will have a more difficulttime proving to employers that they deserve achance to learn and move upward. On-the-joblearning in the new services maybe just as im-portant as in the older manufacturing indus-tries, but the preponderance of mental and so-cial skills over manual skills will lead employers

to place more emphasis on educational creden-tials for entry.

Greater flexibility and improved competitive-ness will prove a two-edged sword in anotherway: some service industries may be able to in-crease their competitiveness only at the expenseof jobs and job opportunities. In others, outputmay grow sufficiently to keep everyone at work,or to create new jobs. But companies that canrespond quickly and effectively to shifts in ex-change rates and world market conditions, theuncertainties of consumer demand, and changesin technology itself, will always be well-placedto prosper in international competition, Tostrengthen its long-term competitive capacity,the United States will have to strengthen its in-stitutional mechanisms for continuing educa-tion and training (ch. 10).

Most of this chapter has focused on indus-tries that have invested heavily in computersand communications, doing so in rather specu-lative fashion, But jobs in the traditional, ter-tiary services—including those least likely tobe automated—will continue to depend in vari-ous ways on the knowledge-based sectors; a na-tion that grows more competitive in high-value-added industries will, all else equal, have higherliving standards and more jobs for everyone,Because of this, one of the most effective, if in-direct, roles for government comes through thecomplex of policies that support and encouragethe development of human capital. This impliesnot only education and training—traditionalresponsibilities of Federal, State, and localgovernments—but aid for the adjustments andtransitions involved when people and compa-nies find themselves moving into new and dif-ferent forms of work organization. The ration-ale is straightforward: by helping ease and guidethe transition to a knowledge-based economy,public policies can contribute to maintainingU.S. employment, to the international competi-tiveness of U.S. firms, and to the Nation’s stand-ard of living.


APPENDIX 8A: RESTRUCTURING AND WORKIn the late 1970s, MetroBank—a medium-sized

competitor in a large urban/suburban market—entered a crisis from which it is still recovering.Burdened with bad international loans, and lockedinto long-term deposit certificates at a time of ris-ing interest rates, the company faced a cash flowsqueeze; it was saved only when a consortium ofother banks lent it capital. When the consortiumpartners insisted that MetroBank restructure itsoperations, the president and executive team werereplaced, many middle managers were fired,branches, subsidiaries and real estate holdings weresold, and liabilities (including the deposit base) werecut in half, A new strategy and new structure meantnew job responsibilities for nearly all the bank’s em-ployees.

New Products, Changing Jobs

Two years into its reorganization, MetroBankfaced the question of formulating a strategy suitedto its altered circumstances. Recent history showedit could no longer be a national player, while sug-gesting a focus on the regional market. Althoughmany commercial clients had withdrawn their ac-counts when the bank was threatened, the retailbase remained loyal. Household customers identi-fied with Metro Bank, a local mainstay for decades.To build on this base in a period of deregulation—with ongoing competitive threats from other finan-cial service firms, including entrants from out-side the banking industry —MetroBank sought tobroaden its range of products and to stress personal-ized service. This high-value-added niche strategyencompassed three main elements:

1. development of 14 new banking products (e.g.,credit cards, Keogh and individual retirementaccounts, money market accounts, money-saver checking accounts, car leasing), thusoffering consumers a portfolio of services;

2. an active selling program at the branch levelintended to increase the number of accountsand products per household; and

3. decentralization of responsibility for commer-cial lending to smaller businesses from themain office to the branches.

Together, these three steps were meant to stabilizethe deposit base by increasing customer reliance

19This appendix is based on a case study prepared for OTAby L. Hirschhorn, Wharton Center for Applied Research, undercontract No. 533-597o. The name of the bank and some detailshave been changed.

on

REORGANIZATION AT METROBANK19

the bank, and to focus lending within Metro-Bank’s market. Senior managers hoped to develop“relationship” banking at the branch level bystrengthening ties with neighborhood businesses.At this time, MetroBank operated about a hundredbranches, and employed 2,000 people.

With the new strategy in place, MetroBank’s em-ployees describe their jobs differently. Branch per-sonnel, including managers, refer to a shift fromoperations to sales, from administration to plan-ning, from order taking to selling. As one branchmanager put it, “Before . . . all you had to do wasto see if there were enough supplies, if procedureswere followed, and get the paperwork done. Youwanted to get people in and out the door. Now . , ,we’re in the retail business. ”

Employee roles changed, particularly for cus-tomer service representatives, branch managers,and area managers:

Customer Service Representatives. Prior to thecrisis, the CSRs (or platform workers, as manyemployees call them) helped customers balancetheir check books and open new accounts, Theywere a cut above secretaries. But MetroBank’ssenior managers believed that the CSRs had tobecome the backbone of the selling effort ineach branch, They were to “cross-sell” prod-ucts—to convince customers coming in for helpto buy other products and services. Metro-Bank’s management created a budget for re-training the platform workers, and instituteda measurement system to assess their sales per-formance.Branch Managers, In the past, branch manag-ers were primarily administrators, evaluatedlargely on their ability to control costs. Sincethe restructuring, they have been asked to gen-erate loans, particularly among businesses withrevenues of less than $10 million. Many of thebranch managers, lacking prior lending experi-ence, have enrolled in the bank’s ongoing creditand lending courses. In addition, branch man-agers must coach and supervise sales effortsby the platform workers.Area Managers. This is a new position, the leastdefined of the three. Each area manager over-sees a group of branches, helping with plan-ning, marketing, and budgeting.

As this summary suggests, MetroBank’s restructur-ing will, if successful, upgrade the jobs and skillsof many employees in its 100-some branches, CSRswill sell rather than simply respond to inquiries,branch managers develop loans on the outside as


well as administer operations on the inside, Em-ployees will have to cope with greater levels of un-certainty as the bank tries to actively shape its mar-kets, rather than taking them as givens.

The Transition

Developing new skills and managing the shiftingrelationships between main office and branches hasmeant a series of adjustments, many of them lessthan smooth.

New Skills

The transition has been a difficult one for the hun-dreds of platform workers expected to do the bulkof the selling and cross-selling of new products thatrange from credit cards to second mortgages forfinancing a child’s education. MetroBank pur-chased two off-the-shelf training courses from a ven-dor; all CSRs took the course dealing with sellingskills, plus a shortened version of a course on fi-nancial counseling. The courses proved no morethan marginally effective.

The platform workers face three sets of difficul-ties: taking authority; managing uncertainty; andplanning their time. Two studies by consulting firmsindicated that CSRs feel out of place selling thebank’s products and services. A shopper’s study (inwhich pseudo-customers came in with unstructuredrequests) found that platform workers could explainbank products effectively, but frequently failed toclose the sale. Interviews with the platform work-ers showed that most still see their work as customerservice-e.g., explaining bank statements—not sales.Almost all complained that they did not have timeto cross-sell; moreover, they tended to see manage-ment pressure for greater selling effort as simplyasking for more work. Striking disparities charac-terize the language of senior managers, who speakof culture change, and the words of CSRs, describ-ing what to them seems a “speed-up. ” Earlier, mostof the platform workers (and other line employees]functioned primarily like clerks and secretaries.While their resistance reflects some of the classictensions between managers and workers, it also re-flects the difficulty of instituting major changes inthe everyday responsibility of a relatively largegroup of employees.

Eventually, the question of career prospects andambitions may prove central. Metro Bank has cutits costs in part by using more part-time tellers anda ‘‘tight platform in many branches. The companyhas continued to view its line workers as expensesrather than assets, while at the same time wanting

them to take on more responsibility. Interviewsshow that few CSRs identify with the bank as awhole. Many like to work close to home. They arenot only reluctant to take on authority and to sell,they don’t want a career that would take them toother branches, One said, “If I wanted to go up, Iwould have been out of this branch a long timeago. . . . All I want is my 10 years and to vest mypension. ” Interviews also indicate that many CSRsstill think of themselves in secretarial terms, rais-ing the question of whether MetroBank is recruit-ing the right kinds of people. The desire to controlcosts leads to hiring of platform workers in the la-bor market for secretaries and clericals. Yet the im-age held by managers suggests someone who wantsauthority and opportunities to rise in the bank.

Technology

Some staff and managers at MetroBank believethat improvements to the computer network usedby the platform workers will help. The redesign ofthe system will proceed in three stages. At first,CSRs will be able to call up information but not en-ter data. In the second stage, they will be able tochange files on-line. Eventually, they will be ableto produce final documents ready for customer sig-nature. Experience at other banks suggests that thiswill help give employees a new perspective on theirjobs. Platform workers will have full and completedata on the customer and on all the bank’s prod-ucts. They will no longer need to call the main of-fice for information. In addition, they will be ableto walk customers through various “what-if” dem-onstrations (e. g., the pay-out on an IRA] as a sellingaid. Eventually, their terminals will be able to dis-play taped presentations on the bank’s products. Asthe terminal becomes an information utility, theplatform workers may come to accept that they arein positions of real responsibility and authority.

Thus far, redesign of the computer network hasnot gone smoothly. Branch and division managershave had trouble working with the systems division.Systems engineers have shown little knowledge of,and little interest in, the development of user--friendly software—nor have they made any effortto learn from the platform workers, or get feedbackfrom them, Most strikingly, the systems division hasfailed to provide training, despite repeated requestsand protests from MetroBank’s line managers.

Managing Uncertainty and Planning Time

Previously, the CSRs simply helped customers asthey entered the branch, Platform workers did not


have to plan who they would see, when and why,for how long. Now, they have been asked to developa sense of priorities — to gage the value of time spentwith each customer, and the value of time spent onservice as opposed to sales. As one branch managerput it, “In an 8-hour day, you can’t wait on one cus-tomer for an hour. I tell my people to develop eachcustomer to the maximum. They should also try toget their operations work done outside of contacthours, ” Some platform workers complain that theyhave too much paperwork, and cannot do the tele-phone selling expected of them after the branchcloses in the afternoon.

The CSRs must also learn to live with uncertainty.The outcome of a sales encounter may not be appar-ent for weeks; success can only be measured byaverages. Psychologically, this means that the plat-form workers must learn to live with failures by bet-ting on future successes. Earlier, the platform work-ers could get a sense of accomplishment simplythrough helping customers as they entered thebranch.

Relationships Between Branches and the Main Office

This aspect of the transition has two dimensions:working relationships between superiors and subor-dinates (which changed with the decentralizationof lending responsibility); and relationships be-tween line and staff. Before restructuring, brancheswere viewed as little more than mail drops for themain office’s lending business. MetroBank’s newstrategy calls for branch managers to spend halftheir time calling on potential loan customers. Tohelp them develop lending and selling skills, thebranch managers were given credit and salesman-ship courses. Senior executives also wanted to shifttheir managers’ attention from expenses to profits,from absolute deposit levels to market share. As onearea manager said “I tell my branch managers, ‘Youare a profit center. ’ We are not telling them, ‘Youmust ring zoo door bells . . ,’ we are saying, ‘tell uswhat you did, what you achieved. ’ “

Thus the rules have changed just as much forlower level managers as for the platform workersthey supervise. One branch manager said, “. , . incommercial lending, you walk out the door notknowing if you have accomplished anything.” Atthe same time, branch managers must now coachtheir own subordinates, rather than simply moni-toring their performance. Formerly, the CSRs andother branch employees had little discretion; theysimply followed standard procedures. Now, the su-pervisors face a more difficult job: helping CSRslearn to sell, and to use the new computer system.

Interviews indicate that few branch managers, asyet, have mastered the coaching process. Indeed,one noted that he needed supervision from his areamanager: “We need someone to take a look at indi-vidual management styles, how you’re doing com-pared to the norms. Not necessarily someone to lookover your shoulder, but someone to go out on a salescall, and sit in on a staff meeting. ”

Planning is also new for the branch managers.MetroBank’s marketing department regularly askseach manager to assess his branch’s performance,evaluate its strengths and weaknesses, and preparea business plan. The head of the market planningunit has expressed considerable disappointmentwith this exercise, reporting that the branchmanagers were not very self-critical, argued thatthey had no real control, and were insensitive tomarket share as a measure of success. The market-ing department has since taken over preparationof business plans for the branches.

The Area Managers

The most striking characteristic of an areamanager’s job is its poor definition, In principle,area managers are in charge of marketing for agroup of 5 to 10 branches. Yet they have no market-ing budgets, One said, “I need permission to spend$2,000 for a party at a senior citizens home. ” Nordo area managers systematically review the per-formance of branches.

The poorly defined role of area managers maysuggest that MetroBank has focused too heavily onthe lower levels, placing most of the burden ofchange on those in the branches. Upper manage-ment may be “leading from behind, ” reluctant tocede real power to those lower in the hierarchy eventhough this is necessary for decentralization towork. whenever top managers delegate authority,they face loss of control, MetroBank’s senior execu-tives may have felt more comfortable delegating tothose at the bottom—whose authority will in anycase be limited to a single branch—rather than tothe area managers.

Such possibilities highlight what MetroBank ex-ecutives call the problem of “creating a new cul-ture. ” Those at all levels speak of the old cultureas a bureaucratic one. They know that success inthe future depends on becoming less bureaucratic,which means authorizing people to act with moreindependence. But the older culture, and its norms,persists. while wanting the platform workers to takemore initiative, some managers think they’re lazy,Meanwhile, ample evidence exists that platformworkers lack the support (e. g., coaching, computertraining) needed to be effective in their new roles.

Ch. 8—Moving Toward a High-Ski/l Economy: Computer Applications and Work Organization in the Services ● 291

Customers

In many high-value-added industries, productsare tailored to fit the client’s needs; the customerhas an intrinsic role in the production process, ineffect co-producing the service. This leads Metro-Bank’s senior managers to speak of relationshipbanking, while for the platform workers it meanseducating the customer. In interviews, many CSRsreflected on customer confusion over the featuresand benefits of IRAs and other new products: “. . . alot of people don’t know anything about money mar-ket accounts ., . people are afraid to use overdraftprotection , , . customers can’t see that if they payoff on their credit cards in time there is no financec h a r g e .

In fact, marketing experiments suggest that manyof Metro Bank’s customers are quite conservative.Few have responded to offers such as free check-ing and free travelers checks as part of packagesincluding other products. Some of MetroBank’sproduct development efforts have failed for reasonsthat executives attribute to customer confusion.Regardless of the reasons for such failures, they

highlight the need for firms in high-value-added in-dustries to develop a good grasp of customer desiresand motivations. In pursuing its high-value-addedstrategy, MetroBank has become dependent on howits consumers see the world, on what they want.

Summary

For MetroBank’s strategy to succeed, its employ-ees, including managers, will need new skills andattitudes. At the most general level, this meanslearning to live with uncertainty. For the bank asa whole, the marketplace is less predictable becauseof deregulation. For individual employees, uncer-tainty (and anxiety] comes with the need to sell newretail products and bring in loan business. By giv-ing the platform workers more immediate accessto information, and by increasing their apparentcompetence in the eyes of customers, the redesignedcomputer system may help resolve some of the ten-sions. But several years after restructuring, Metro-Bank cannot yet feel confident in its new strategy.

Documents

Chapter 8 Moving Toward a High-Skill Economy