academy ol manaqemeni executive, 1999, vol. 13. no. 4

Academy ol ManaqemenI Executive, 1999, Vol. 13. No. 4

Surprising but true: Half thedecisions in organizations fail

Paui C. Nutt

Executive OverviewHalf the decisions in organizations fail Studies of 356 decisions in medium to large

organizations in the U.S. and Canada reveal that these failures can be traced tomanagers who impose solutions, limit the search for alternatives, and use power toimplement their plans. Managers who make the need for action clear at the outset, setobjectives, carry out an unrestricted search for solutions, and get key people toparticipate are more apt to be successful. Tactics prone to fail were used in (wo of everythree decisions that were studied. >

Why Decisions Fail

Half the decisions made in organizations fail.' Thisis a major finding from studies of decision makingthat I have conducted over the past two decades.^This dramatic failure rate cannot be explained byconventional wisdom. For instance, failure doesnot generally stem from things beyond a manag-er's control: draconian regulations imposed bygovernment, unexpected budget cuts by higher-ups, or loss of market share because of fickle cus-tomers. Although failures can occur when regula-tions run up costs, when budget flexibility is lost,and when customer preferences shift and wreck amarketing plan, the tactics that managers use aremore important. Studies of decision making that Ihave conducted over a 20-year period trace failureto managers who employ poor tactics.^

Why are tactics prone to fail used so often? Mywork suggests three reasons: Some tactics with agood track record are commonly known, but un-commonly practiced. Nearly everyone surveyed isaware of participation and its ability to coax ac-ceptance, but participation is used in just one offive decisions. Another reason is that decisionmakers take short cuts when they feel time pres-sure. As pressure appears to mount, reason givesway to such homilies as, "Why invent the wheel?"The practices of a respected organization are thencopied, which is rationalized as timely and prag-matic. But such short cuts often lead to unantici-pated difficulties and delays as steps are taken toconvince people that the company's interests, notyours, are being served. A third reason for failure

is subleties. Telling people you want to lower costsis much more powerful than finding the root causeof the cost problem. Managers who are drawn toproblem solving fail to see that problems promptblame. Telling subordinates what's wanted liber-ates them to look for answers. Finding problemsalerts subordinates to take defensive action. En-ergy is directed away from finding answers to pro-tecting their backs and their interests.

To find out why decisions go wrong, I began myresearch by collecting real decisions in real orga-nizations, made by real people. Getting close to theaction uncovered tactics and allowed me to see adecision's result and its consequences. Connectingoutcomes to tactics provided a telling appraisal ofthe effectiveness of the tactics employed by man-agers.

Telling suboTdinates what's wantedliberates them to look /or answers.Finding problems alerts subordinates totake defensive action.

Managers can recall their successes and fail-ures, but seldom subject them to systematic anal-ysis. Erroneous associations between tactics andtheir results result and managers may discard per-fectly good ways of making decisions and continueusing others with poor track records. This articiesummarizes some of the things I have learned fromsuch analyses.^ Lessons drawn from this work areoffered to suggest what to do and what to avoid to

75

76 Academy of Management Executive November

increase your chances of making successful deci-sions.

Inventory of Day-to-Day Managerial Decisions

The cases are based on 356 decisions made bysenior managers in medium to large organiza-tions across the United States and Canada. (SeeTables 1 and 2). They are the sorts of decisionsmanagers face day-to-day about new products,equipment purchases, staffing, pricing, market-ing, controls, planning, personnel policy, con-struction, and customer service. About a quarterwere made in public agencies, about half in pri-vate sector profit-making companies, with theremainder in third-sector organizations. Publicorganizations are made up of governmentalagencies funded by tax dollars. Private organi-zations are for-profit firms that offer productsand services paid for by consumers. Third-sectororganizations include private, not-for-profit or-ganizations, such as the bulk of U.S. hospitals,charities, symphony orchestras, and profes-sional societies. The results do not vary muchbetween the different types of decisions and or-ganizations.

Decision-making practices were collected by in-terviewing key participants, including the personresponsible for the decision."* The steps managersfollowed were uncovered and classified into tac-tics that are used to set directions, find solutions.

Table 1Decision Types, Organizations, and Participants

in the Cases

Table 2Illustrations of Organizational Decisions

Organizations Decisions

Characteristics Number Percent

Decision Types

Service or ProductTechnology (e.g., equipment)Internal operation (e.g., inventory control)Control systems (e.g., accounting software)Reorganizations (e.g., mergers, restructuring)Personnel policy (e.g., benefit packages)Markets (e.g., pricing, advertising)

Total

OrganizationsThird-sector (e.g., private nonprofits)Private (e.g., for-profit companies)Public (e.g., governmental agencies)

Total

ParticipantsExecutives (CEO, COO, CFO)Middle managers

Total

94706954351915

356

14812684

356

231123

262019151054

100

413523

100

6535

356 100

University HospitalBoss Laboratories

Florida MedicaidOhio DNRU.S. Air ForceNASAVeteransCity oi ColumbusOhio DOTMichigan HealthPublic School SystemU.S. NavyMcDonald'sKorean Tire Co.Nationwide InsuranceAllied Van LinesMarshall FieldsBank OneFifth-third BankG.E.National City Corp.LennoxMead PaperAnthony ThomasDelcoCompuServeBethlehem Steel

BattelieToyotaLarge cityThe Limited, Inc.American ElectricGeneral MotorsKorean AirlinesHuntington BankAmerican Telephone400-bed acute care

hospitalMcDonnell-Douglas575-bed acute care

hospitalDunning LathmpTwo small fast food

restaurants1000-bed university

hospital343-bed acute care

hospitalCity health service500-bed acute care

hospital1000-bed acute care

hospital250-bed rural acute care

hospitalLane Bryant. Inc.For-profit journal

abstractingA large companyDelco ElectronicsHertz-Penske RentalNCR

Scheduling operating roomMarketing infant formula to

developing nationsFraud management systemSupporting wildlife programsProcurementDecompression serviceRestructuringDivesting hospice programBudget system revampingDisposing of contaminated cattleRedesigning curriculumRadar developmentNew location and designMarketing in South AmericaComputer system capacityPricing servicesNew product lineSelling Visa cardsDropping Saturday serviceMRP systemPrivate label credit cardRecycle toxic wasteCost cutting systemNew productTariff managementNew on-line serviceScheduling blast furnace

maintenanceContract biddingIncreasing salesCreation oi retirement centerPurchase an information systemCAD/CAM systemRobotic assemblersStaff cutbackBilling and collection proceduresMarketing planAdd a lithotripsy service

TQM teamsSelect a radiation treatment simulator

Modify bonus policyInitiate a joint venture

Purchase a magnetic resonanceimager

Add a pulmonary treatmentprogram

Locate a halfway house for the deaiAdd an open-heart program

Provide a helicopter transportation

Create a detox unit

Intimate apparelDevise a reference library

Create a marketing programInventory control systemCustomer service systemCash flow management

1999 NuH 77

and implement the solution.̂ '̂ Decision outcomeswere identified and success measured, linking tac-tics to the success measures with statistical tech-niques.

The primary indicator of success was whethera decision was put to use. Changes in conven-tional wisdom, awareness, enlightenment, or at-tempts to legitimize did not count as successes.'^A management information system (MIS) wascalled a failure if the organization continued touse the old system, and a merger was a successif it was completed. Each decision was followedfor two years to determine changes in use thatcan occur with time. During this two-year period,some decisions unraveled, being used only inpart, or not at all. With this information, deci-sions were classified in the study as successfulor not according to their long-term use (decisionssustained for two years) and degree of use (de-cisions still in full use after two years). Indicatorsof each decision's value and the time required tocarry it out were also obtained.^

Organizational Decision Making

A process describing what managers do as theymake a decision is shown in Figure 1.̂ The processstages identify what managers worry about as de-cisions are being made. Different people empha-size different things. Some stress finding a work-able idea and deemphasize direction setting.Others emphasize implementation and still othersconsider all the stages and give them equalweight. The path shown by arrows in Figure 1 isoften called a rational process, in which all thestages are considered in an orderly manner. Atactic indicates how managers go about uncover-ing the things that are called for by a stage: setdirections with an objective to indicate what'swanted or copy what others do to get an idea.

The process gets started when a signal capturessomeone's attention. Signals identify somethinginside an organization, such as inefficient opera-tions or loss of legitimacy, or outside it, such as aninnovation by a competitor or a loss of customers.

DiagnosisSignals

Questions about organizational eifectiveness

Information Gkithering

Findings

HoStop -«-

ActionYes Performance realized

(e.g., use)

Establish Direction Motivating concerns Implementation

What's wanted A solution

Identify Options Evaluation

Some ideasPlan Development

Alternatives

Note: Stages discussed in this article are in bold boxes.

HGURE 1Decision-making Process


that calls into question organizational practices.These questions prompt information gathering,such as studies by staff, examining industry re-ports, and talking with stakeholders or trustedpeers. A motivating concern emerges when theinformation gathered made the signals seem im-portant. Concerns then take shape, such as copingwith changes in customer tastes or expectations, ornew computer technologies. When such a concerndraws attention to an unsatisfactory situation, itprompts action.

Managers respond by working their way throughthe process stages, selecting tactics along the way.The orderly path in Figure 1 is seldom followed,because managers often jump to conclusions andthen try to implement the solution they reached.The bias for action causes them to limit theirsearch, consider too few alternatives, and pay toolittle attention to people who are affected, not re-alizing that decisions fail for just these reasons.

To concentrate on the most important lessons,the three stages in which tactics have the greatestinfluence on success are considered. The tacticsemployed by managers to set directions, uncoveroptions, and implement their plans identify pro-cess types that are more and less successful. Theless successful imposition process type uncovers asolution early on and does not consider any othersolution ideas. The more effective discovery pro-cess type establishes directions and identifies op-tions as separate activities. Options are developedin response to the directions established. Somediscovery processes deal with implementationfirst. In others, implementation comes after a solu-tion is uncovered. A decision in which implemen-tation concerns are addressed early on is moresuccessful. Table 3 identifies the key features ofeach tactic and provides an illustration. Table 4summarizes the frequency with which each tacticis employed and the long-term and full-use ratethat results.

Tactics Used to Set Direction

Managers in the cases established a direction byusing one of the following four tactics, which arenamed to indicate how directions are set.

Generating Ideas

The most common way of establishing a direction,employed in 37 percent of the cases, is with anidea. Managers find what seems to be a usefulidea in the signals motivating action and fashion itinto a ready-made solution. For example, a com-pany approached an air-conditioning manufac-

turer with an idea for a solar heat pump during anenergy crisis. A license was offered to sell the heatpump in exchange for developing a heat wheel, akey part of the solar heat pump's design. The air-conditioning firm agreed and spent eight years inan attempt to devise a heat wheel with both mois-ture retention and durability, but failed to find asuitable construction material. The CEO made noattempt to move away from the design problem, toredesign the solar heat pump without a heatwheel, or to look for a better way to provide energy-efficient home heating and cooling. The projectwas abandoned by the CEO only after the declin-ing cost of energy removed the incentive to findnew sources of energy.

The idea tactic prompts managers to focus on asingle solution, as did the air conditioning firm'sCEO. They saw the idea as a pragmatic way totake decisive action, and made no effort to findanother option. They believed that decisivenessallowed them to quickly manage possible threatsthat could spin out of control. Speedy action isalways favored, even in situations that have noreal time pressure. However, managers in thesecases often struggled to verify the virtues of theiridea, to coax support from others, and had to re-peatedly modify the idea to make it workable.Commitment became a trap that often producedfailure. Indeed, solutions derived in this way werefully used in only 42 percent of the cases, and only56 percent of those were used for the entire twoyears. Managers became trapped by perceptionsof sunk cost, by perceived threats in admittingfailure, and by reluctance to start over. Many orga-nizations reward only success. This makes manag-ers fear the appearance of failure, admitting thatmoney has been spent without producing anythingof value and that it will take longer than antici-pated to get results.

Many organizations reward only success.This makes managers fear theappearance of failure, admitting thatmoney has been spent without producinganything of value and that it will takelonger than anticipated to get results.

Problem Solving

In more than 26 percent of the cases, managersdefined a problem and then analyzed its distinc-tive features with the hope of quickly uncoveringclues that would suggest a remedy. For example,the director and his staff of the Ohio Department of

1999 Nutt 79

Claims saw a growing backlog of social securitybenefit appeals and called for a change in han-dling procedures. Analysis led to a pooling ideathat grouped similar claims for mass handling.However, the backlog analysis failed to focus onthe reason for the growing number of claims. Afterthe backlog grew to the point that claims took ayear to process, the director stumbled on a loop-hole in the legislation that inadvertently easedeligibility requirements. The director made thelegislature aware of the oversight and the loopholewas closed. In the meantime, the agency was sub-jected to constant criticism and legal action for itsslow, error-prone claims management. As the casesuggests, defining a problem is a familiar way formanagers to initiate decision making.'° Managerswant to find out what is wrong and fix it quickly.The all too frequent result is a hasty problem def-inition that proves to be misleading. Symptoms areanalyzed while more important concerns are ig-nored.

Problem solving is no more successful than theready-made solution. Only 44 percent of the solu-tions that resulted from problem solving were fullyused, and only 55 percent of them were sustainedfor as long as two years. The low sustainabilityoccurred because it became increasingly obviousover time that more important things had beenoverlooked. Fortunately, more successful tacticsfor establishing directions can be found in thecases.

Setting Objectives

Managers set an objective to guide decision mak-ing in 30 percent of the cases. The objective indi-cates the results that a manager wants to realize,such as lower cost or increased market share. Thisapproach gives everyone involved considerablefreedom to search for solution options. Such man-agers are open to anything that would provide thedesired result. Setting objectives results in a 70-percent sustained use and a 58-percent full use,making it one of the better ways to set directions.

Objectives are commonly known, but uncom-monly practiced because managers often have abias toward action and fear being seen as indeci-sive. Action-oriented managers see objectives asan academic exercise. Identifying desired resultsseems obvious, and devoting time to somethingthought to be obvious is irritating to the action-oriented manager. Such managers stress the needto move forward, and have little patience with ob-jective-setting sessions. Also, many managerswant to be seen as decisive. This creates an arti-ficial pressure for action. The pressure takes sev-

eral forms. Managers are expected to guaranteethey will fix a problem the moment a concernemerges. Saying what will be done as soon as aconcern materializes makes the manager seem tobe on top of things. Authorizing a study with ob-jectives is often sneered at by the press and othersin oversight roles. Such reactions make it difficultto champion an orderly process that clearly artic-ulates desired results with an objective and waitfor a solution. Managers who would prefer to fol-low such a path are often pressured by higher upsor people in oversight roles to grab the first ideathat pops up. People creating such pressure act asif all concerns have an immediate solution. Evenwhen managers know that making decisions inthis way is foolhardy, the pressure for a quick fixoften wins out.

Ironically, setting objectives has just the oppo-site effect. Objectives liberate people to searchwidely for solutions. This lowers the chance offailure, and therefore of criticism. For example, ahospital CEO responded to a threat by Blue Cross,an insurance carrier, to cut its reimbursementrates. The hospital's proposed service charges ne-gotiated by the hospital CEO biannually and for-malized as a contract with Blue Cross, had beenrejected. Blue Cross claimed that the hospital wasoverstaffed. To respond to this threat, the hospitalCEO identified a cost reduction target and let de-partmental managers determine how they wouldmake the necessary cuts. The cost reduction targetdirected the search for ways to reduce labor cost. Asuccessful rate negotiation with Blue Cross re-sulted when the CEO indicated the cuts to bemade.

Setting objectives can prompt failure if the ob-jective is too demanding. Such an objective sur-passes what can be achieved, given the time andresources available, and may panic and frustrateparticipants." The chance of success improveswhen a realistic objective is set. At this point, youmay be wondering about the stretch objectivefound in the writings on total quality management{TQM} and reengineering.^^ ^ y reading of this lit-erature suggests that the cases cited in TQM andreengineering are mostly anecdotal and seem tohave one common element—^poorly used re-sources. For example, long distance carriers suchas Sprint, AT&T, and MCI were not satisfied withBell Atlantic's time to hook up new customers. Thir-teen hand-offs and seven information systemscaused repeated delays as people coordinatedwith one another or waited for replies, producing a15- to 30-day wait period. The objective selected,the 10 hours of actual work, was hardly a stretch.Redeployed resources found in unneeded or inef-

80 Academy ol Management Executive November

ficient procedures, not stretch objectives, explainthis result. On the other hand, the continued use ofstretch objectives in stressed companies that areshort on resources may be a prime cause of thewidely-reported erosion of morale and growth ofapathy in U.S. companies.'^

/ntervening in the Process

Intervention was the least frequently used tactic,observed in just seven percent of the cases, yetwas the most successful. These decisions weresustained for two years in 96 percent of the casesand were still in use after two years in 92 percent oithe cases. To intervene, a manager demonstratesthe imperative to act by comparing current perfor-mance to norms that discredited it. This demon-stration justifies the new norm by benchmarkingthe performance levels of respected organizations.

Intervention was the least frequentlyused tactic, observed in just sevenpercent of the cases, yet was the mostsuccessful.

Managers compare current performance to the newnorm and offer some ways to improve perfor-mance. Considerable time is spent networkingwith key people, explaining where the new normscame from, documenting performance, and callingattention to ideas that could work. The perfor-mance gap that results is used by these managersto make claims about needs and opportunities.These claims for action provide direction for thedecision-making effort, indicating what is wantedas an outcome.

Managers often assume that the concern thatmotivated them to act is obvious to others. Infor-mants in my cases pointed out that key peopleoften have no knowledge of the reasons motivatinga decision-making effort or disagree about its im-portance. This causes managers considerablegrief as they try to reassure doubting or belligerentstakeholders, gobbling up time and delaying ac-tion. Key players often remain skeptical as justifi-cations are belatedly offered, and speculate aboutthe manager's hidden agenda, even when thereisn't one. Managers who apply the interventiontactic are able to alter such perceptions and in-crease their chance of success.

The steps taken by managers to carry out inter-vention can be illustrated by a corporate reorgani-zation plan. The company's board of directors hadbecome frustrated with its scope of responsibili-

ties, which became their motivating concern. Theboard members called attention to the need forchange by identifying responsibilities that are typ-ical for comparable boards. A comparison of theseresponsibilities suggested that the board had toomuch involvement in day-to-day operations. Theoversight responsibilities in comparable organiza-tions indicated what needed to be changed andhow the board's fiduciary responsibilities could becarried out with less involvement in day-to-dayoperations. The chairperson used this informationto network with key stakeholders to explain theneed for change.

Lessons for Establishing Direction

• The rates of use and success records of the di-rection-setting tactics fail to match. The mostsuccessful tactics are infrequently used, and theleast successful frequently used. Avoiding ideaand problem-solving tactics and using objectiveor intervention tactics will improve the chance ofsuccess.

• Problem-solving tactics fail because the searchfor solutions is narrowed and defensiveness isevoked. For example, when a problem with ad-vertising appears to exist, search is limited toadvertising solutions, ruling out solutions thatdeal with anything else. A manager attemptingto cope with an advertising problem is oftenunable to look beyond this problem toward op-portunities that others see as useful. Poor resultsalso stem from the defensive behavior promptedby a problem analysis. Listing problems makespeople defensive because they see themselvesas potentially accountable. This prompts themto maneuver to avoid becoming scapegoats anddraws energy from the decision effort. However,problem listing is seductive and managers usethis approach in one-quarter of their decision-making efforts.

• Intervention and setting objectives are effectivebecause they encourage learning and develop-ment.''' When the direction set by a managerrapidly narrows to a specific problem, or dis-places a solution, little learning can occur. Byopening up the decision process to new possi-bilities, managers are more apt to recognize thevalue of new ideas and move away from stereo-typed responses and traditional ways of acting.

Managers find it difficult to set directions usingthe best tactics because they are drawn to powerand shy away from ambiguity and uncertainty.Powerful managers often impose their ideas on thedecision-making effort when seductive-looking so-

1999 Nutt Bl

lutions crop up. Seizing such opportunities isviewed as a pragmatic way to take decisive action.Telling people how to make a decision is a time-honored way to appear decisive. However, impos-ing a solution frequently leads to failure. Manag-ers may prefer to make sense of a decision after thefact,'̂ and to combine chance with opportunity.'^Preconceived solutions and limited searches foroptions are recipes for failure.

When managers impose an answer, they createa misleading clarity that sweeps aside importantsources of ambiguity and uncertainty. Considerthe heat pump case. After the solar heat pumpconcept had been accepted by the company's CEO,no one was prepared to challenge its merit. Theexpectation of an increase in sales was never ver-ified or even discussed with anyone in the organi-zation as the decision process unfolded. Decisionswith an idea direction seldom go beyond that idea.Managers become fixated with the idea and do notask reframing questions, such as, "How do we finda product that complements our air-conditioningbusiness?" or, "How do we increase sales?" Subor-dinates are reluctant to carry bad news that theidea does not work, as eight years of repeatedfailure in the heat pump case indicate.

Managers do not know what they want untilthey see what they can get.'' Having a specificanswer sweeps away ambiguity and uncer-tainty. At first this is comforting, but it subse-quently limits a manager's ability to see attrac-tive options. Furthermore, beginning with animposed solution is rash when a manager haslittle understanding of what is needed. Instead ofseeking understanding, such managers explorethe virtues of an idea and focus on the reactionsof key people to it. The more insistent managersbecome, the more time they must spend defend-ing the idea. Managers who did not impose asolution, and took steps to justify the need to actusing intervention tactics cut the duration of thedecision process by 300 percent. Even better,managing the politics of the situation with thesesteps greatly improves the prospect of decisionsuccess. Ideas can be helpful when used to illus-trate the benefits of new practices, but thesebenefits will be lost if the options under consid-eration are limited to the initial set of ideas.Treating ideas as possibilities, as in the inter-vention tactic, improves the prospects of success.

Tactics Used to Identify Options

In the 133 cases that used discovery processes, anobjective or an intervention was followed bybenchmarking, searching, or designing to uncover

options. The key features of these tactics can befound in Table 3 and their frequency of occurrenceand success rates in Table 4.

I

Benchmarking the Best Practices of Others

Managers often benchmark practices that are be-ing used by respected organizations to uncover asolution. Practices that seem desirable are ex-pected to meet an objective or the needs suggestedby an intervention. These managers said that theycopied the practices of others because it seemedpragmatic and would cut costs. As one of the par-ticipating managers noted: "Why reinvent thewheel when someone else may have done it foryou?" This kind of logic made benchmarking themost widely used approach for identifying options.

A single-benchmark tactic cropped up in sevenpercent of the cases, and had both sustained andfull use rates of 59 percent. Managers using thistactic copy the practices of a single organization orwork unit thought to have high prestige, and thentailor these practices to fit their needs. In one case,a manager hired someone who had successfullyinstalled a desired system for a leading competi-tor. Along with hiring a solution, single bench-marks are derived from site visits made by man-agers and from descriptions published inperiodicals and books.

Failures stemmed from difficulties that arose asthe transported idea was modified to fit the needsof its new user. For example, the CEO of the Uni-versity of Minnesota's Medical Center visited theUniversity of Iowa's Medical Center to study itscorporate and governance structure. The visitorenumerated the functions of trustees, the preroga-tives of its medical staff, and mechanisms of ac-countability for state and federal funds. Unfortu-nately, the Iowa Medical Center has a state-widemonopoly over tertiary care treatment for the poor,giving it complete control over substantialamounts of revenue. The structure that evolved tofit this situation offers no guidance in forming agovernance structure for the University of Minne-sota, with its more competitive environment.

Some managers used the more sophisticatedapproach of integrated benchmarking. This tac-tic was observed in six percent of the cases andhad a good record of success with 71-percentsustained and 78-percent fully used decisions.Managers who use integrated benchmarking ex-amined the practices of several organizations orwork groups, identifying the best features fromeach. An amalgamation of these practices pro-duced the solution. For example, before purchas-ing a major piece of equipment, several organi-


Table 3Decision-Making Stages and Their Key Tactics

Stages/Tactics Key Features Illustration

Establish DirectionImposition process

types• Idea

• Problem solving

Discovery processtypes

• Objectives

Intervention

Identify OptionsImposition process

(ypes• Existing solution

Discovery processtypes

• BenchmarkingSingleIntegrated

• SearchingSingle

Multiple

• Designing

Implement Decisions

• Intervention(revisited)

• Participation

Token

DelegatedI

Complete

Comprehensive

• Persuasion

• Edict

Impose a ready-made solution on the decision-making effort

Infer a solution by problem analysis

Set objectives to guide solution development

Find new norms that dramatize (he need foraction. Suggest needs and opportunities toguide solution development.

Explore the idea

Adapting the ideas of othersSingle sourceAmalgamate ideas from multiple sources

Aggressive and overt searchOne search cycle

Learning from sequential searches

Custom-made solutions are sought

Show stakeholders that performance hasimproved and that performance now meetsagreed upon norms

A task force made up of stakeholders formedto make recommendations:solution framing with partial participation

solution specification with partialparticipation

solution framing with full participation

solution specification with full participation

Attempt to sell a solution by demonstrating itsbenefits

Issue a directive

A company attempted to develop a solar heat pumpdevised by another firm.

The Ohio Department ol Claims analyzed its claimbacklog to find reasons for its growth.

A hospital identified a cost reduction target and letdepartments determine how to make the cuts to meetthe target.

A board of directors called attention to its scope ofresponsibilities by comparing them to those of boardsin comparable companies. The need to reduce theseresponsibilities was identified by !he atypical degreeof the board's involvement with the firm's operations.

Management shows the benefits ol a solar heat pump orthe remedy for claim backlogs.

A material management system was obtained by hiringsomeone who had developed such a system.

Before purchasing equipment, several organizationswere visited to see how the equipment was used. Thebest features of practice at each site were combined toform a plan.

An RFP was used by a firm to select among computervendors.

A recursive search was carried out by a firm for anauditor, using what was learned in a past search toidentify requirements for the next search cycle.

A new system was designed for the records, scheduling,and billing patients in a university hospital clinic,without reference to existing systems.

Hospital trustees were shown that the high cost ol bumcare had been covered by endowments,reimbursements, and cheap resident manpower.

A committee made up of departmental representativeswas formed to identify concerns.

A committee made up of departmental representativeswas formed to uncover concerns and recommendsolutions.

Surveys of people near a proposed prison site wereconducted to identify their concerns.

(not observed)

A CEO asked the head of an EDP department to makerecommendations for a new computer and used thearguments to make a presentation to the board ofdirectors.

A materials management department head issued amemo indicating new responsibilities for people whowere affected by the plan or must carry it out.

1999 Nutt 83

Stage/Tactic

Establish Directionsimposition process types• Idea• Problem Solving

Discovery process types• Objectives• Intervention

TOTAL

Identify optionsImposition process fypes• Existing solution

Discovery process types• Benchmarking

SingleIntegrated

• SearchingSingleMultiple

• Designing

TOTAL

Implement Decisions• Intervention {revisited)• Participation

TokenDelegatedCompleteComprehensive

• Persuasion• Edict

TOTAL

Success

Cases

13192

10726

356

223

46(25)(21)43(34)(9)44

356

2663(9)

m(20)(0)

133132

344

Table 4of Decision-Making Tactics

Percent ofDecisions Studied

s»307

m

63

13

mm12

m

100

T18(3)

(10)(6)(0)3ft3S

100

Sustained UseRate

565S

7096

N/a

55

5971

6310063

N/a

96807080

1000

58S3

62

Full UseRate

4244

5892

N/a

41

5978

5110053

N/a

92738777950

4735

50

Notes: Decisions that used multiple tactics were not included. N/a = not applicable.

zations were visited to determine the types ofequipment in use and to extract the best featuresof procedures for staffing and other aspects ofusing the equipment. A hybrid of procedures andequipment with the best match to local condi-tions suggests the solution.

Four reasons accounted for the infrequent use ofthe integrated benchmarking tactic. First, many ofthe managers seem unaware of this tactic or itsbenefits. Second, managers seem to know littleabout systems analysis and other techniques re-quired to create a systems synthesis,'^ Third, somemanagers have limited access to relevant prac-tices because high status organizations are lockedin fierce competition with the manager's organiza-tion. Finally, perceived time pressure begins tomount as decision-making reaches the idea stage,often creating artificial pressure to adopt the firstworkable idea that is uncovered.'^

Searching for Solutions

In 12 percent of the cases managers used searchaids, such as a request for proposal, or RFP, to findprepackaged solutions from vendors or consult-ants. Search efforts can be either single or multi-ple. Managers who feel that they are aware ofstandards by which to judge a proposed optioncarry out a single search. Others conduct a multi-ple search, such as the manager who searchedrepeatedly to learn about the financial analysispackages offered by consultants. Such managersaccumulated several competing packages andcompared them to discern their features and capa-bilities. With this knowledge, a new RFP called fora system with features known to be available andneeded by the organization. In a single search, theRFP was written with available information, andlearning and choosing were done at the same time.


The multiple search tactic was used in three per-cent of the cases with no failures. The singlesearch tactic was used in 10 percent of the cases; 51percent of these decisions were fully used and 63percent were sustained for at least two years.

Managers who experience failure are often un-aware of what is available. Managers who do notknow what they do not know or want are apt to beduped. Vendors often sell a system whether or notit fits the user's needs. Huge expenditures aremade for data processing systems with features anorganization has no use for; the firm must paysomeone else to write software to make the systemmeet its needs. The multiple search tactic in-creases the prospect of finding off-the-shelf ideasthat fit your needs.

Managers fail to use the multiple search tacticbecause they are unaware of its benefits, or feeltoo much time pressure. However, managers with aworkable option, hoping to find a better one, candefend a multiple search by citing the reducedprospect of failure associated with it.

Designing Options

Managers apply design to develop innovative op-tions in response to the need or objective guidingthe decision-making effort. They used design todevise marketing programs, controls, products,and services with original features—distinct fromthose found in the practices of others. Design tac-tics were found in 12 percent of the cases andproduce 63 percent sustained decisions and 53 per-cent fully used decisions.

Managers were reluctant to use design becauseit seemed quite risky, compared with the bench-marking or the searching tactics. The motivationsto be pragmatic are stronger than the urges to beinnovative. This is unfortunate because failure hasless to do with the risk of innovation than with howdesign is carried out. Poor results often stem from

The motivations to be pragmatic arestronger than the urges to be innovative.This is unfortunate because failure hasless to do with the risk of innovation thanwith how design is carried out.

poor practice. Managers are rarely schooled in de-sign and few have access to people who are. Inaddition, many of the subordinates who are askedto come up with an innovative solution have noidea how to proceed. Consultants using the designtactic are even more prone to failure than in-house

staff because they design systems based on thewants of their client instead of their needs. Thedemands for repeat business force consultants tobe very sensitive to what their clients say theywant and make it difficult to criticize these wantsfor fear of offending clients.

Lessons lor Identifying Options

• The ineffectiveness of the single benchmark andsingle search tactics stem from several relatedfactors. Imposing an idea seems to be pragmaticbecause it's fast and decisive. However, mystudies suggest that managers who used singlebenchmarks often spent considerable time try-ing to make the idea work, as they did in thecases in which a solution was imposed at theoutset. The single benchmark provided a solu-tion often selected in haste, with little reflection,and then required considerable tailoring lateron to get it to work. Furthermore, some of themanagers with a ready-made solution had hid-den motives. Choice opportunities enticed somemanagers to bring out a self-serving idea thatwas unrelated to any of the organization's con-cerns. Resources were mobilized to justify thesolution, which kept the people in the organiza-tion from looking into other possibilities.

• Single searches opened up the search processbut allowed little opportunity for learning, com-pared with the integrated benchmarking andmultiple search tactics. Integrated benchmark-ing and multiple search prompted managers ina decision-making role to engage in a process ofreflection in which possibilities were carefullystudied. This investment paid dividends by im-proving success. Time was actually saved be-cause fewer repairs were necessary to fix solu-tions gone awry.

• Design is controversial. Design produced goodresults under certain conditions, such as for im-portant decisions and when multiple alterna-tives were sought. Design was less successfulunder other conditions, and never reached thesuccess level of the multiple search and inte-grated benchmarking tactics. This mixed en-dorsement of design is also found in the litera-ture, where the practice of design and itsfrequency of use are disputed. Some argue thatdesign should be used more often, which wouldincrease the number of innovation attempts inorganizations.2° Others view innovation as highrisk,2i which suggests that, even with the userate of 12 percent, design may occur more fre-quently than is desirable. However, you can im-prove the prospects of success when good de-

1999 Nutt 85

sign practices are followed that stress creativity.When good practice is followed, the prospect ofsuccess improves. To increase the rate of inno-vation in organizations, it may be necessary totrain managers and their staffs in both creativitytechniques and in approaches that help organi-zations absorb new ideas.

• Nearly every discussion of decision makingcalls for developing multiple options. Multipleoptions allow managers in a decision-makingrole to combine the best features of options tomake a superior one, and to make a comparisonwith a preferred action to demonstrate its merits.In my research, managers used multiple optionsin less than 20 percent of their decisions. Whenmultiple options were developed, success ratesjumped from 56 percent to 70 percent. Multipleoptions provide another way to increase yourchance of success.

Tactics Used to Implement Decision Stages

Four tactics are used by managers to pave the wayfor installing a preferred course of action. I callthem intervention, participation, persuasion, andedict.

/nf erven fion

Implementation by intervention occurred in sevenpercent of the decisions studied, with 96 percentsustained and 92 percent fully used. This makesintervention the most successful, and the least fre-quently used, approach to implementation. Thesuccess of intervention cuts across organizationallevels. Middle managers had 90 percent of theirdecisions sustained, and 75 percent fully used,compared with 100-percent full use when interven-tion was used by top managers. The top manager'sposition helps to guarantee success. Middle man-agers must carefully work through a process ofauthorization and justification of the proposed newnorms. This takes more time than some are willingto devote. Although their prospects of full adoptionare less, middle managers can dramatically in-crease their prospects of success by using inter-vention.

Intervention creates the need for change in theminds of key people, identifying and justifyingnew performance norms. Showing how a compara-ble organization is able to operate with lower costcreates new cost expectations, suggesting a realopportunity to make a positive change. After asolution was found, these managers intervenedagain to facilitate implementation by showinghow performance could be improved. For example.

hospital trustees in one of the cases were wary ofoffering high-cost services, such as burn and cor-onary care, because third party payers like Medic-aid and insurance carriers do not cover the full costof treatment. To overcome this objection, a hospitalCEO demonstrated how the cost of burn care couldbe covered by a variety of sources—endowments,reimbursements, and cheap resident manpower.The CEO presented the trustees with a demonstra-tion of financial feasibility and a description ofhow other hospitals had enhanced their image,and thus endowments, with this service. A candi-date for the burn care supervisor position showedhow resident recruitment had been improved athospitals after burn service had been added, sug-gesting that future resident recruitment effortscould be harmed without a burn care service. TheCEO reported back to the trustees, providing dataon image, resident recruitment, and cost after apilot test of the unit.

Parficipation

Managers started other decision-making efforts bycreating task forces with key individuals as mem-bers. Authority was delegated to the task force tooversee important aspects of the effort. Managersemployed participation in 18 percent of the casesand had 80 percent of their decisions sustainedand 73 percent fully used.

Managers said that they were aware of the valueof participation, but found it difficult to use. This isanother example of commonly known, but uncom-monly practiced, activities. Managers reportedavoiding participation because of its time require-ments and the seeming loss of control that results.They had little knowledge of the high failure rateof other tactics, or how to deal with obstacles thatarise when involving people to make a decision.Another explanation can be found in subtletiesthat stem from the paradox of control, which arenot as widely known. Managers who give up con-trol through participation actually get more con-trol. People are more apt to ask for help when theyneed it and be candid about the decision situationwhen asked to participate in the decision-makingeffort. Unilateral action will close off this type ofinformation.

Managers using participation form task forcesthat have differences in the degree of involvementand roles for its members. This suggests four typesof participation. When a few people affected by thedecision are given limited involvement, token par-ticipation results. Only a subset of the affectedindividuals are consulted and given little to do. Forexample, the dean of a college of business formed


an executive committee to comment on his policyideas for the college. Two faculty were selected torepresent the more than 100 faculty members, lo-cated in six departments, who were expected to beaware of and represent the concerns of their col-leagues—a difficult task. The dean used the exec-utive committee as a sounding board to test hisideas and seldom asked for ideas from the com-mittee. Token participation can also result whensurveys are conducted to sample user needs orproblems. Choosing a site for a regional headquar-ters was carried out in this way in one of the casesto determine what people wanted and would ac-cept.

Delegated participation also limits the involve-ment of stakeholders, but asks the task force to domore, such as offer a solution. This type of partic-ipation is used in strategic planning. A strategicplanning group is formed that draws its membersfrom the organization's board and its top manage-ment, and is expected to create a plan. Managerswho use complete participation involve all thestakeholders as members of a task force and callon the group to comment on some aspect of adecision. For instance, in one of the cases, ail ac-countants in a firm's accounting department wereasked to state their needs before a financial anal-ysis package was purchased. Comprehensive partic-ipation requires that all stakeholders be involved inboth finding concerns and offering solutions.

Delegated participation is often used, token andcomplete participation rarely used, and compre-hensive participation never used. (See Table 4). Intheory, all affected parties can be asked to identifyand select among proposals, but comprehensiveparticipation was not observed in my research.Questions about cost and unpredictability arisewhen decisions are delegated to this extent. Man-agers seemed to be unaware that token and dele-gated participation have a lower rate of successthan does complete participation. Token participa-tion had a 70-percent sustained use rate and 67-percent full use. Delegated participation worksbetter than token participation, with 80-percentsustained and 77-percent complete use rates.When task force members are given an importantassignment, success is more likely. Complete par-ticipation is rarely employed but leads to decisionsthat are adopted in almost all cases. Wide involve-ment is even more important than a meaningfultask. Participation failures are linked to low in-volvement. As the proportion of participants to allaffected parties falls, the failure rate for participa-tion increases. The power of co-optation, enticingpeople who participate to go along, is difficult toexport. A few enthusiastic participants are unable

to sway people who have vested interests or aresuspicious about the manager's motives.

Persuasion

Managers use persuasion in nearly 40 percent of thecases to implement decisions. These managerscalled on experts to search out options offered byvendors or being used by competitors, or to devisenovel options, carefully evaluating the merits of theproposed solution. Managers combined the rationalarguments provided by experts with salesmanshipto convince people to go along with a decision. De-spite its frequent use, the persuasion tactic has sig-nificantly lower success than the participation andintervention tactics, with a 56-percent sustained use,and only a 47-percent full use rate.

Managers mistakenly saw persuasion as low risk.If the expert could convince them, they believed theycould convince others. These managers saw persua-sion as more successful than the success measuresindicated. Failure occurred when the experts' argu-ments were unable to sway people. Gathering doc-umentation to support the merits of a decision wasoften done at the expense of gaining the acceptanceof key people affected by it. Unmanaged social andpolitical concerns, such as job security and vestedinterests, scuttled many of these decisions.

Edicts

Some managers use their power to issue a direc-tive that announces a decision. A memorandum iswritten, job training conducted, or an administra-tor hired to carry out actions called for by thedecision. This is done without consulting with peo-ple who have stakes in the changes the decisionwould bring. For example, a new pricing policy foremergency room services was announced by amemorandum sent to all hospital employees andmembers of the medical staff. The memo explainedthe new policy and when it would go into effect.Edicts were observed in 40 percent of the cases andhad the highest failure rate. When implementationwas attempted by edict, 53 percent of the decisionswere sustained for two years, and only 35 percentwere fully used.

Managers know that they must rely on theirpower to issue an edict, but believe that their pre-rogatives and the need for timely action make thisjustifiable. They seem unaware of the high rate offailure of an edict. Failure can be traced to under-estimating that amount of power needed becausepeople resist the appearance of being forced tocomply. To use an edict you must draw on yoursocial credit, that store of goodwill built up by

1999 Nutt 87

honest dealings and positive accomplishments.When you are carrying out an edict, your socialcredit is traded for quick action. Repeated use ofedicts depletes the store of social credit, openingthe way to sabotage, token compliance, delays,and outright refusals to comply.^^ The turmoil thatresults seems to outweigh the decision's benefitsso these decisions are often withdrawn.

Managers who routinely use edicts develop areputation of being heavy-handed and put theirdecisions in jeopardy. Attention is directed awayfrom the merits of the decision and toward themanner in which implementation is attempted.Considerable expense, time, and effort are ex-pended trying to overcome the resistance provokedby an edict. More often than not, these effortsfailed, no matter what the decision's merits. Forexample, one firm tried to force internal use of anMIS system that had been developed by a subsid-iary. The system had no external buyers, makingthe targeted in-house users see it as suspect. De-partment managers resented being forced to adoptan MIS of questionable value that could adverselyaffect their performance. When asked to partici-pate in a pilot effort, they simply refused to coop-erate. The CEO of the firm entered the fray. Incen-tives, in the form of budget supplements, wereoffered if a department would operate the new MISalong with its current system, and compare theresults. The department designated to work withthe subsidiary sabotaged the effort by continuingto rely on its old information system and by pro-viding erroneous information to the new system,causing it to malfunction. The malfunctions werethen cited by the department as evidence that thenew MIS design was faulty, ultimately forcing thefirm to abandon its attempts to use the subsidiary'sMIS system.

Lessons for Implementing Decisions

• Managers should become more involved in de-cision making. When a manager takes chargeand creates an environment in which change isjustified and understood, implementation is al-most always successful.^^ However, managersdistanced themselves from implementation bydelegating important tasks in 93 percent of thedecisions studied.

• Substitute intervention for edicts and persuasion.The success of implementation by intervention ex-tends to all decision-making efforts, regardless offactors such as initial resistance to change, thescope of change called for, decision importance,resources available to support the decision effort,urgency, and the decision maker's level in the

organization.2^ This suggests that substituting in-tervention for edicts and persuasion could nearlydouble decision-making success.

• Involve people affected by a decision.^^ The hu-man relations movement has long determinedthat implementation success improves whenkey people are involved. But there is an impor-tant qualification suggested by my work. Repre-senting some stakeholders can help, but thepower of co-optation does not extend to peoplewho are not involved in decision-making activ-ities. Managers often limit involvement, withjust a few representatives of a large group ofstakeholders placed on a task force. Or the taskforce is limited to commenting on the decision,or other relatively minor tasks. Leverage is lostwhen participation is curtailed in these ways.Complete participation is rarely observed, per-haps because it can be difficult or impossible toinvolve large numbers of scattered stakehold-ers. For example, decisions that affect largenumbers of people, such as locating prisons orairports, can never hope to involve all the inter-ested parties. Nevertheless, managers will bemore successful if they expand the role of theirtask forces as much as possible to involve asmany of the stakeholders as is feasible, andkeep others informed about what is being con-sidered. If intervention cannot be used becauseof the press of other activities, participation is aviable substitute. When using participation in-stead of intervention, managers should antici-pate a drop in the prospects for success fromwell above 90 percent to about 70 percent.

• The origins of implementation by persuasioncan be traced to the operations research move-ment.̂ E Advocates of a sales approach arguethat success depends on anticipating and coun-tering objections to proposed changes. Expertsdetermine what should be done and devise ra-tional arguments to support their views. Orga-nizations like the National Science Foundation,the Ford Foundation, and the National Institutesof Health use this approach in their peer reviewprocedures; Experts select the grants and con-tracts to be funded by the agency and offer jus-tifications for their choices. The controversialOperations Research Society of America guide-lines, published in 1971, called for experts tovigorously sell ideas that they find desirable.When experts recommend and managers react,implementation becomes an exercise in sales-manship. In 40 percent of the cases, managers ina decision-making role delegate to experts inthis manner. Such managers give the expertbroad responsibility, but withhold approval un-

Academy of Management Executive November

til the expert can demonstrate value. This forcesthe expert to concentrate on providing compel-ling arguments that support the preferred courseof action. Gathering arguments that support adecision occurs at the expense of gaining theacceptance of stakeholders, and accounts forpersuasion's 50-percent failure rate.Issuing edicts calls for reward, legitimate, expert,informational, or referent power.̂ ''' Reward poweris the authority to offer incentives or remove irri-tants. To apply this type of power, managers mustcome up with an incentive that entices people togo along with their decision. Legitimate power isbased on a manager's right to take action. Expertpower flows from a track record of success and areputation for candid and honest dealings. Expertpower requires both credibility and trust, but in-formational power, in which a manager offers in-sights or cites anecdotes, requires only that themanager have the trust of other organizationalmembers. Referent power stems from the individ-uals seen as charismatic, such as Ross Perot whofounded EDS, or Lee Iacocca, who saved ChryslerCorporation from bankruptcy. Managers draw onone or more of these power bases when using anedict. For example, legitimate power is appliedwhen a decision maker employs his or her prerog-ative to make a decision. When carrying out anedict, managers trade their social credit for actionand risk failure if their power proves insufficient.The frequent use of power strains an organizationand gradually drains the manager's store of socialcredit. Eventually, social credit disappears andevery decision becomes a test of will. Morale de-teriorates and the organization sacrifices the cre-ativity of its members for their compliance. Thisaccounts for the high rate of failure associatedwith issuing edicts.

Successful Decision Making

Managers can improve their chances of makingsuccessful decisions. These suggestions stem frommy key findings about tactics, important relatedpractices to follow, and tips on how to manage thedecision-making effort. These suggestions call onyou to practice what you know, resist pressures fora quick fix, accept uncertainty and ambiguity, andrecognize subtleties in what works and whatdoesn't. Following these suggestions has minimalcost when compared to the cost of a failed deci-sion.

• Personally manage your decision-making pro-cesses. The prospects of success improve whenyou take charge. Delegation to experts or to peo-

ple who are expected to champion your ideasmay give you time for other things, but willmake success much less likely.

• Search for understanding. Signals that captureyour attention can be symptoms of other con-cerns that are misleading, or more urgent thanimportant. Careful probing can provide a win-dow that can open up on a landscape with use-ful insights into what needs fixing. The timespent in reflecting on what is at issue can payhandsome dividends. A deeper understandingof the issues meriting your attention providesbetter direction as well as a defense for thecourse of action selected.

• Establish your direction with an interventionand an objective. Intervention establishes therationale for action. An objective that indicatesthe desired outcome opens up a search to newideas. An open search pays dividends by reduc-ing the chance of failure.

• Stress idea creation and implementation. A de-cision-making process should guide thinkingabout action and taking action. Many decisionmakers see the importance of one but not theother, being biased toward idea development ormanaging the politics of the situation. There isno substitute for clear thinking or for diplomaticaction. Both thoughtful idea development andadroit idea promotion are essential.

There is no substitute for clear thinkingor for diplomatic action. Both thoughtfulidea development and adroit ideapromotion are essential.

• Identify more than one option. The considerationof several competing options improves decisionresults. The discarded options are not wasted.They help confirm the value of a preferredcourse of action and frequently offer ways toimprove it. Employ one of the more effectiveoption development tactics to do this. Considerdeveloping an option with integrated bench-marking, with cyclical search, and with design.This opens up your search process to a variety ofideas from different sources. The best one or acombination of the best features of each suggestsa solution that improves your chance of success.

• Deal with barriers to action. Implementation tac-tics must address social and political barriers toaction to be successful. Intervention is the bestway to manage the social and political barriersthat can block a decision. Participation can berecommended when using intervention would

1999 Nutt 89

draw your attention away from other, more im-portant activities. Avoid edicts and persuasion,even when a decision seems urgent.

Endnotes' This finding from my research over the past two decades

was highlighted in The Wall Street journal. 1996. Business Bul-letin. July 25.

'' These conclusions stem from a series ol studies I have done,including: 1984. Types ol organizational decision process, Ad-ministrative Science Quarterly. 29 (3), 414-450; 1986. The tacticsof implementation. Academy of Managemenf/ournai, 29 (2}.23O-261: 1987. Identifying and appraising how managers installstrategic changes. Strategic Management Journal, 8 (1),!-14;1989. Selecting tactics to implement strategic plans, SfrategicManagement huinal. 10, 145-161: 1993. The formulation pro-cesses and tactics used in organizational decision making,Organizafion Science, 4 (2), 226-251; and 1993. The identificationof solution ideas during organizational decision making, Man-agement Science. 39 (9). 1071-1085.

^This article draws on some key ideas from more than 30studies reported in the literature.

* The approach to interview people and proiile decisions wasadapted from the work of Patton, M. P.. 1990. Qualitative eval-uation and research methods, Newbury Park, Calif.: Sage, andDenzin, N. K., 1989. The research act: A theoretical introductionto sociological methods, Englewood Cliifs, N.J.; Prentice Hall:and Lincoln, Y. S. & Guba, E. G. 1985, WafuraJislic inquiry,Beverly Hills: Sage.

'' The steps taken to discover tactics were: a.) A top executive inthe participating organization suggests a current decision andprovides documents describing it. b.) The decision maker and twoothers familiar with the decision and responsible for carrying itout are recruited for interviewing. In most cases, the contact per-son offered a decision he or she made and became the primaryinformant, c.) Interviews were conducted by the author, asking thedecision maker and one other informant, independently, to spellout the sequence ol steps taken to make the decision, d.) A writtensummary was prepared by me. e.) The summary was presented toeach informant, (.) The summary was modified by informants untilit agreed with their recall of events, g.) At an additional interview,the decision maker was asked to amend the summary to accountlor dillerences in documents and the other informant's account ofwhat happened. Only cases that had agreement and sufficientdetail to understand what was done were included in the data-base of decisions, h,) The two secondary iniormants were asked toindicate if the decision was used, to indicate how long it took, andto evaluate it. Two years from the date ol first implementationthese indicators were collected again, i.) I prepared a one-fwgesummary that listed the key decision steps taken, j.) Patterns inthese steps were identified to uncover tactics by sorting cases.putting each in a pile that used similar steps or tactics untilgroupings emerged that seemed mutually exclusive: putting codenumbers on back of each summary and repeating the sort to findambiguous classifications; creating new category or modifying anexisting one to resolve ambiguities: repeating until full agreementwas obtained (intrarater reliability 100 percent), and having acolleague repeat the sorting (interrater reliability 90 percent), k.)Link tactics and measures of success using statistical techniques.

^ The explanation-building ideas of Yin were used to uncovertactics. See Yin. R.. 1993. Applications of case study research:Design and methods. Beverly Hills: Sage.

' This distinction between symbolic and instrumental used canbe found in Pelz, D.C.. 1978, Some expanded perspectives on use ofsocial science in public policy, in Yinger, M. & Cutler, S. J. (Eds.)

Major social issues: A multidisciplinary view, New York: Free Press(346-357), and Beyer. J. M, & Trice. H. M., 1982, The utilizationprocess: A conceptual framework and synthesis for empirical find-ings. Administrative Science QuarferJy, 27, 591-622.

^The measures of use increased as indicators of decisionvalue increased, and tended to decline as indicators of durationincreased. Thus, successful decisions were put to use, had highvalue, and were finished in a timely manner.

^ Similar stages have emerged in a wide variety of studies.Sources considered in my studies included Damanpour. F.. 1991,Organizational innovation: A Meta-analysis of determinantsand moderators, Academy of Management joutnal, 34(3),555-590: Dunn, W. N., 1981. Public policy analysis, Englewood Cliffs,N,I.: Prentice Hall; Hage, J., & Aiken, H., 1970. Soctai change incomplex organizafions, New York: Random House; Nutt. P. C,1992. Managing planned change. New York: Macmillan: andMarch, J. G., 1994, A primer on decision making: How decisionshappen. New York: The Free Press.

'"An interesting discussion of the impact ot symptomatic andmisleading problem directions can be found in Kolb, D. A., 1982.Problem management: Learning from experience, in Srivasta, S.(Ed.), The executive mind. San Francisco: Jossey-Bass, 109-143.

" The difficulties that stem from panic-driven decision mak-ing are discussed in Janis, I., 1989. CruciaJ decisions, New York:The Free Press.

'̂ For some illustrations, see Linden, R, M., 1994. The seamiessorganization, San Francisco: Jossey-Bass.

'^Kelley. R,, 1992. The power ol followership. New York:Doubled ay.

'•• Pettigrew, A., 1995. The awaking giant: Continuity andchange at ICI. Oxford: Blackwell.

'̂ Weick, K., 1979. The social psychology of organizing, 2ndedition. Reading, Mass: Addison-Wesley.

'̂ The notion of chance decision making is discussed in Cohen,M. D., March. ]. P., & Olsen, ]. P., 1976, A garbage can model oforganizational choice. Administrative Science Quarterly, 17,1-25.

"See Wildavsky, A., 1979. Speaking truth to power, Boston:Little Brown,

^̂ A useful discussion of systems synthesis can be found inWeinberg, G. 8f Weinberg. D., 1979. On the desigrj of stablesystems. New York: Wiley-Interscience.

'̂ March and Simon describe the motivation to use blandalternatives in their classic work: March, J. G., & Simon, H. A.,1958. Organizad'ons. New York: Wiley.

™ Nadler, G., & Hibino, S. 1990. Breakthrough thinking, Prima,Rocklin, CA.

^' Another view of innovation can be found in March, J. G.,Footnotes to organizational change, Adminisfrafive ScienceQuarterly, 26,4(1981), 56S-577.

^̂ ' A useful discussion of the resistance provoked by edictscan be found in Bardack, E.. 1977. The implementation game.Cambridge: MIT Press.

^̂ See Dalton, G. W., Influence and organizational change. InDalton, G, W,, Lawrence, P., & Greiner, L. (Eds). 1970. Organiza-(ionai Change and Development. Homewood. IL: Irwin.

'̂' Nutt. P. C, 1998. Leverage, resistance, and the success ofimplementation approaches. Journal of Management Studies.35(2),1998, 213-240.

^̂ See, for example, Coch, L. & French, ]. R. P., Jr. 1948. Over-coming resistance to change. Human Belations, 1,512-532.

^ See. for example. Churchman, C. W. 1975. Theories of imple-mentation. In Schultz, R. L. & Slevin, D. P. (Eds), /mpJemenfingoperations flesearch/Managemenf Science. New York: Elsevier.

^' See, for example, the classic work of French, J., & Raven, B.,1959. The bases of social power. In D. Cartwright (Ed.), Studies ir:social power: 150-167. Ann Arbor: Institute for Social Research.


Paul C. Nutt is a professor ofmanagement sciences in theFisher College of Business atThe Ohio State University, withappointments in Public Policyand Management and Indus-trial Engineering, He receivedhis Ph.D. Degree at the Univer-sity of Wisconsin-Madison anda BS and MS from the Univer-sity of Michigan, all in indus-trial engineering. He has writ-ten over 100 articles and sevenbooks. Confac(: nu[(,/@osu.edLi.

academy ol manaqemeni executive, 1999, vol. 13. no. 4

Documents