work redesign: eight obstacles and opportunities

8/10/2019 WORK REDESIGN: EIGHT obstacles and opportunities

1/24


2/24

368 H UMANRESOURCE MANAGEMENT, Winter 2005

the practicality of the topic. In addition, cur-rent trends in human resource managementresearch that consider strategic HR (Delery &Shaw, 2001) and human capital manage-ment (Lepak & Snell, 1999) can be aided byconsidering theoretical and practical impli-cations of job-design research. For example,Delery and Shaw (2001) and Tsui, Pearce,Porter, and Tripoli (1997) argue that manystrategic HR policies should differ by job in

the organization. Decisions sur-rounding such differential poli-cies would benefit from knowl-edge of the design of those jobs.For example, a decision to imple-ment a gainsharing compensa-tion system for a subgroup of jobsnecessitates an understanding of

the interdependencies amongthose jobs. Similarly, the humancapital management perspective,which concerns the effective uti-lization of human capabilities,would clearly benefit from en-hanced knowledge of job design,in part because job design has im-plications for what people do andhow effectively they do it.

HR professionals today canprofit greatly from the knowledge

gained through these years of re-search and practice with job de-

sign. There are several clear relationships be-tween characteristics of work and employeereactions that can guide efforts to simultane-ously maximize efficiency and satisfaction inthe workplace. Several challenges, or obsta-cles, remain, however, for practitioners at-tempting to implement work-design changes.

These challenges emerge for diverse rea-sons, such as conflicting constituent needs,the complexity of organizations, and the prac-

tical realities of the workplace. For example,work designed according to mechanisticallyoriented principles will be radically differentfrom work designed according to motivation-ally oriented principles. The seemingly irrec-oncilable trade-offs between the two ap-proaches represent an obstacle to research andpractice because they suggest a dichotomy:work can be either efficient or satisfying.

The purpose of this article is to improveunderstanding of these eight obstacles andto provide direction for managing them. Asthese obstacles are formidable and defy sim-ple solutions, it is not our intention to com-pletely solve each of them. Rather, we hopeto clarify sufficiently the issues involved andto aid practitioners in making informed andrational work-design decisions that comple-ment their particular situation.

Obstacle #1:Work Design InfluencesMultiple Outcomes

The first obstacle lies in recognizing thatwork design influences multiple outcomes. Infact, different scientific disciplines have pro-duced several distinct approaches to job de-

sign and research. Further complicating mat-ters is that each approach has beenconducted relatively independently of theothers. The interdisciplinary job-design per-spective of Campion (1988, 1989; Campion& Thayer, 1985) highlights this fact and sug-gests that there are at least four basic ap-proaches to work design, each focused on adistinct set of outcomes. These four ap-proaches are labeled mechanistic, motiva-tional, perceptual, and biological (Table I).Practitioners need to consider all four ap-

proaches when redesigning work. Failure tosimultaneously acknowledge differences inthe purpose, primary outcomes, and findingsof each approach has impeded the progress ofwork-design research in defining comprehen-sive and practical models to aid practitioners.If the number and type of outcomes consid-ered are constrained to those found within agiven model, practitioners will face obstaclesimplementing effective work-design changes.

The first approach, the mechanisticmodel, is grounded in classical industrial en-

gineering research (Barnes, 1980; Gilbreth,1911; Taylor, 1911). This model evolvedlargely to manage pressures for efficiencythat arose during the Industrial Revolution.Simplification, specialization, and repetitionof work are the central tenets of the model.Advocates of this approach believe it can in-crease efficiency, make staffing easier, reducetraining costs, and lower compensation

There are several

clear relationships

between

characteristics of

work and employee

reactions that can

guide efforts to

simultaneously

maximize efficiency

and satisfaction in

the workplace.


3/24

Work Redesign: Eight Obstacles and Opportunities 369

requirements. This approach is similar tocontrol-oriented HR strategies (Arthur, 1992,1994) and compliance-based HR configura-

tions (Lepak & Snell, 1999).Proceeding primarily from research in

organizational psychology (Hackman &Oldham, 1980; Herzberg, 1966), the motiva-tional model evolved in response to job dis-satisfaction, the deskilling of industrial jobs,and the alienation of workers that resultedfrom the overapplication of the mechanisticmodel. The approach usually provides jobenriching recommendations such as in-creasing the variety of tasks performed orthe autonomy with which they are exe-

cuted. The intended benefits of this modelinclude increased job satisfaction, intrinsicmotivation, retention, and customer service.This work has clear connections with com-mitment-oriented HR strategies (Arthur,1992, 1994) and HR configurations (Lepak &Snell, 1999).

Based on human factors and experimen-tal psychology research (Fogel, 1967;

McCormick, 1976; Meister, 1971), thepercep-tual model arose from increases in technolog-ical complexity and a shift in many jobs

from manually performing work to operat-ing and monitoring. This approach is prima-rily concerned with reducing the informa-tion-processing requirements of work inorder to reduce the likelihood of errors, acci-dents, and mental overload.

Emerging from ergonomics and medicalsciences research (Astrand & Rodahl, 1977;Grandjean, 1980), the biological modelsought to alleviate physical stresses ofwork. Reductions in physical requirementsand environmental stressors and increased

consideration of postural factors are com-mon recommendations. Taking these fac-tors into account when designing jobs canreduce physical discomfort, physical stress,and fatigue.

As mentioned, each of these approachestends to be studied within a single disciplineand focuses primarily on the benefits forsolving one particular problem. It is clear,

Illustrative

Model Discipline Base Recommendations Typical Benefits Typical Costs

Mechanistic Industrial Specialization Efficiency Decreased

Engineering Simplification Easier staffing satisfaction

Repetition Reduced training Decreased

motivation

Motivational Organizational Variety Satisfaction Training

Psychology Autonomy Intrinsic motivation Errors

Participation Retention Stress

Customer service

Perceptual Human Factors Reduce information- Reduced errors Boredom

Experimental processing Fewer accidents Monotony

Psychology requirements Less mental

overload

Biological Ergonomics Reduce physical Physical comfort Financial costs

Medical Sciences requirements Reduced physical InactivityReduce stress

environmental Reduced fatigue

stressors

Note. These models are drawn from research in multiple disciplines that is exemplified by Taylor (1911) for mechanistic, Hackman and Old-

ham (1980) for motivational, Meister (1971) for perceptual, and Grandjean (1980) for biological. Recent evidence for the benefits and costs

is contained in Campion (1988; 1989), Campion and Berger (1990), Campion and McClelland (1991; 1993) and Campion and Thayer (1985).

T A B L E I Interdisciplinary Perspective on Job Design and Associated Benefits and Costs


4/24


however, that each of these approaches is as-sociated with certain costs, although thecosts were not clearly recognized until inter-disciplinary research was conducted that di-rectly compared the models (e.g., Campion,1988; Campion & Thayer, 1985). The costsrepresent the loss of benefits that wouldhave been attained if an alternative modelhad been chosen.

For example, the costs associ-ated with the motivational model,such as increased training require-ments, likelihood of errors, andon-the-job stress, parallel the ben-efits of the mechanistic and per-ceptual models (Table I). In a sim-ilar fashion, designing workaccording to the perceptual model

can result in the undesirable out-comes of boredom and monot-ony, two benefits of the motiva-tional model. Finally, thebiological model is independentof the others but still involvescosts in the form of financial ex-penses to modify technology andwork conditions, as well as workerinactivity that may result if too

many of the jobs physical requirements areremoved. When practitioners conduct work

redesign within one or only some of these ap-proaches and ignore others, the costs are typ-ically unrecognized. This omission may im-pede successful work-design interventions.

The obstacle is particularly problematicto HR practitioners because each of the keyvariables typically is important to organiza-tions. Managers do not have the luxury ofmaximizing job satisfaction at the expenseof efficiency. Similarly, neglecting ergonom-ics to focus on efficiency is not an acceptableoption. Thus, in dealing with this obstacle,

HR managers and researchers should con-sider several key issues.

First, practitioners must ask the ques-tions Do the costs always occur? or Underwhat conditions are the costs more or lesslikely to occur? to determine which out-come variables are of greatest relevance totheir situation. Organizational culture andindividual differences are two sample vari-

ables that a practitioner could account forwhen making decisions about importantoutcomes.

Second, attention should be paid to linksbetween the appropriate job-design ap-proach and the HR strategy of the organiza-tion. For example, the mechanistic approachmight fit better with a control-oriented HRstrategy, while the motivational approachmight complement a commitment-orientedHR strategy (Arthur, 1992, 1994).

Third, current research on strategic HRhas distinguished between core jobs andnoncore jobs, with core jobs being more crit-ical to the core competencies of the organi-zation (Delery & Shaw, 2001). Decisionsabout how to work and job-redesign workmay very well depend on whether a job is

considered to be core to the strategic aims ofthe organization. These considerations willhelp assure that design of work in the organ-ization adequately fits the organizationalculture and strategy, precursors to a success-ful intervention.

Obstacle #2: Trade-offs betweenDifferent Work-Design Approaches

Implied in the previous discussion, a secondmajor obstacle is the inherent tension be-

tween the various models. That is, changesrecommended from each discipline, aimedat improving its specific outcomes, tend tobe incompatible, or even in direct conflictwith, changes recommended from anotherwork-design model. Recent research has ac-knowledged the trade-offs between differentjob-design approaches and under what cir-cumstances the trade-offs are more likely tooccur (Campion, 1988; Campion & McClel-land, 1993; Edwards, Scully, & Brtek, 1999,2000). In addition, these studies have inves-

tigated whether or not trade-offs can be min-imized if jobs are redesigned (Morgeson &Campion, 2002).

The most prominent example is found inthe apparent trade-offs between the mecha-nistic and motivational models of work de-sign (Campion, 1988; Campion & McClel-land, 1991, 1993; Campion & Thayer, 1985).This trade-off arises from the fact that many

Decisions about

how to work and

job-redesign work

may very well

depend on whether

a job is considered

to be core to the

strategic aims of the

organization.


5/24


recommendations from the motivationalmodel increase satisfaction but reduce effi-ciency, whereas the recommendations fromthe mechanistic model increase efficiencybut reduce satisfaction. This tension createsconflict for practitioners who wish to cap-ture the benefits of the mechanistic modelwithout incurring its costs. One manufactur-ing company implemented lean productionto improve quality and standardize workflow. They found that the motivational prop-erties of the jobs, such as autonomy, skill uti-lization, and participation in decision mak-ing, all declined. This implementation alsoresulted in reduced organizational commit-ment, role breadth self-efficacy, and in-creased job depression (Parker, 2003), thusdemonstrating the trade-offs that may occur

when redesigning work.Researchers in this field have been un-able to provide clear direction to practition-ers wanting to maximize multiple outcomesthat span disciplinary boundaries, such as at-taining both efficient and satisfying work.Recent work-redesign research suggests anumber of approaches practitioners can usein response to this obstacle. However, themost appropriate approach will dependupon a host of factors, including the strate-gic objectives of the work design.

The compromise approach involves a directjudgment about the outcomes that are cho-sen as the focus of the work-redesign inter-vention. The desired outcomes then dictatethe job-design model used. This approach isuseful because all parties are aware of thetrade-offs and difficulty of simultaneouslymaximizing all outcomes. Managers makeinformed decisions based upon the out-comes they value most and the costs theywill accept. For example, in Campion andMcClelland (1991), the motivational model

was used in redesigning the work in order toincrease job satisfaction and customer serv-ice, knowing that there would be no gain(and maybe even losses) in efficiency andtraining times.

The level-separation approach involves de-signing different levels of the organizationusing different models. For example, organi-zational structures can be designed accord-

ing to the mechanistic model, and jobswithin departments can be designed basedupon the motivational model. In this way,basic efficiencies are built into the flow ofthe work, yet individual jobs are satisfying.This raises a question for practitioners interms of whether the approaches should beviewed as hierarchical, with the mechanisticapproach more appropriate for higher levels.

The sequential approach to reconcilingconflicts may involve applying one modelfirst, followed by the other. For example, themechanistic model could be used to firstmake the jobs more efficient, andthen the motivational modelcould be applied to make the jobsmore satisfying. Alternatively, themotivational model could be

used to create motivating work,such as a work team having re-sponsibility for an entire product,and then individual jobs could bemade as efficient as possible, suchas having each person in the teamperform a specialized task. How-ever, practitioners should con-sider whether jobs require someminimum level of efficiency inorder to be motivating, because inefficientjobs often are viewed as frustrating. Sequen-

tial strategies also highlight two features ofvirtually all job-design activities: there isequifinality (i.e., multiple configurations cansometimes produce the desired outcomes)and the process is iterative (i.e., improve-ments in jobs often proceed in many smallsteps).

The synthesis approach focuses on specify-ing areas in which gains can be made basedon one model without sacrificing the othermodels. It is sometimes possible to reapsome of the benefits of one model without

incurring large costs in terms of another, ifcareful attention is paid to the specificchanges made. For example, costs of the mo-tivational model are more likely if mentaldemands are increased (e.g., increased au-tonomy and skill utilization), but less likelywhen other changes are made (e.g., increas-ing feedback and social support; Campion &McClelland, 1993), because increased mental

Managers make

informed decisions

based upon the

outcomes they value

most and the costs

they will accept.


6/24


demands create costs in terms of the out-comes of the mechanistic approach (e.g.,training times and compensation require-ments). Practitioners should investigatewhich work-design principles in each modelcreate costs and which do not.

It is also possible that trade-offs that existat the individual level are less apparent at theteam level. The team approach involves usingteam-based design to simultaneously reap ef-ficiency, cost savings, and motivational ben-efits. For example, team members are able tolearn from each other via cross-training,

thereby lowering training costsand increasing workforce flexibil-ity. Team designs may also in-crease motivation through in-creases in social facilitation, skill

variety, task significance, andfeedback (Campion, Medsker, &Higgs, 1993). In addition, work-load sharing by team memberscan help avoid downtime. On theother hand, practitioners shouldbe advised that teams may incurgreater total costs as a result ofwell-known group process losses(e.g., social loafing, conformity).Therefore, further inquiry issorely needed on this issue.

There are several well-publi-cized approaches to work design that may bepopular because they represent hybrids ofmultiple models and thus may help practi-tioners reduce or balance the trade-offs. Totalquality management (TQM) is one strategythat attempts to combine elements of boththe mechanistic and motivational models ofjob design. The general focus of TQM is tocontinually improve processes and productquality (Deming, 1986).

TQMs focus on work processes reflects a

mechanistic orientation. Management bydata (i.e., the importance of feedback), how-ever, is a central construct in motivational,as well as mechanistic, models of work de-sign. Continuous learning is the same as anorientation to learning and developmentthat is inherently motivational, as is the useof cross-functional teams and quality circles.Deming also highlighted the importance of

leadership, communication, and training(DeVor, Chang, & Sutherland, 1992). TQMappears to have withstood the fad stage, per-haps because it balances competing perspec-tives in work-design theory and, thus, offersa viable option for practitioners.

The principles of reengineeringalso reflectconcepts from both the motivational andmechanistic models of job design. Reengi-neering is primarily a mechanistic approachto job design because of its focus on creatingefficiency of methods, facilities, materials,and work flows (Hammer & Champy, 1993).However, reengineering also has componentsof the motivational approach. For example,combining jobs, creating process teams, andincreasing the complexity of jobs all areclearly linked to principles of the motiva-

tional model. Likewise, reducing the numberof checks and controls, performing processesin their natural order, and flattening the or-ganization, all seem related to the motiva-tional model. The long-term survival of thishybrid remains to be seen, however, becauseof its frequent association with downsizingand uncertain record of productivity im-provement (Hammer & Stanton, 1995).

The socio-technical systems approach seeksto improve productivity and satisfaction byconsidering both technological systems

(mechanistic model) and human systems(motivational model) in designing work sothat the two can be jointly optimized (Cum-mings, 1978). One of the primary contribu-tions of this perspective is the use of au-tonomous work teams to accomplish work.Other motivational principles of this ap-proach include increasing task variety, creat-ing a meaningful pattern of tasks that are re-lated to a whole task, establishing optimumwork-cycle length, enabling discretion in per-formance standards and feedback, enlarging

jobs to include boundary tasks, and in-creasing skill levels (Trist & Bamforth, 1951).Socio-technical systems theory has a rela-tively long history, but its key principles havenot been completely tested and validated(i.e., joint optimization and controlling vari-ance at its source; Morgeson & Campion,2003), something we view as warranting fur-ther investigation by practitioners.

Total quality

management is one

strategy that

attempts to combine

elements of both the

mechanistic and

motivational models

of job design.


7/24


This obstacle also can be problematic forpractitioners who need to achieve desiredoutcomes from the redesign process. Practi-tioners may well be faced with the task of re-designing work in order to achieve both effi-ciency and satisfaction. Research offersseveral suggestions for practitioners facedwith this obstacle (Morgeson & Campion,2002). First, practitioners should consideradopting an interdisciplinary perspective onwork design, a factor also indicated in thefirst obstacle. Explicitly acknowledging mul-tiple work-design models will enable practi-tioners to focus job redesign on specific as-pects of jobs and minimize inherenttrade-offs.

Second, practitioners should specify thedesired outcomes of the redesign process.

Most redesigns are aimed at achieving just ahandful of important outcomes and theremay be several options for design thatachieve these outcomes while avoiding trade-offs. Finally, the principle of joint optimiza-tion from socio-technical systems theory sug-gests that the key to minimizing trade-offs isbalance in work design (Morgeson & Cam-pion, 2002). Thus, practitioners should con-sider a variety of approaches to work designin order to reduce needed trade-offs.

Obstacle #3: Difficulty in Choosingan Appropriate Unit of Analysis

A pragmatic issue in applying job-designprinciples is choosing a unit of analysis. Un-fortunately, no guidelines exist for determin-ing the proper unit of analysis for a given sit-uation. This lack of guidance is an obstacleto creating unified theories of job design andto making practical changes to jobs. For ex-ample, most job-design efforts focus on theproperties that jobs should possess, but

then create change at the task level. This dis-connect between the theory and the applica-tion makes it even more difficult for practi-tioners to conduct a successful job design.

Practitioners face this obstacle when re-designing jobs for specific purposes. For ex-ample, companies may centralize activitiesfrom numerous geographical locations, withthe goal of consolidating activities and gain-

ing efficiencies. In this case, analysis con-ducted at the job level may make it difficultto recognize potential opportunities for con-solidation. The analysis should be conductedat a different level in order to consolidateand gain efficiencies. Due to these chal-lenges, a practically useful and theoreticallymeaningful unit of analysis is needed.

We propose four possible levels of analy-sis for practitioners to utilize when redesign-ing work. These choices include jobs, duties,tasks, and task clusters. Jobs represent thehighest level of analysis and can be definedas a group of duties performed bya single individual. The next levelof analysis consists of duties,which are composed of multipletasks that form a major portion of

the work performed. The lowestlevel of analysis is tasks, whichare typically defined as discretework activities (Harvey, 1991; Mc-Cormick, 1979). Finally, we pro-pose a fourth level of analysis, thetask cluster, as an intermediarylevel between tasks and duties.

Given the distinct advantagesand disadvantages associated withusing each unit of analysis, practitionersshould be advised that the choice of unit used

can influence the effectiveness of a work-de-sign intervention. For example, researchersstudying the well-being of nurses divided jobsinto four integrated task categories. This divi-sion allowed practitioners to see which inter-ventions most improved the nurses well-being, since the intervention differeddepending on the task (Le Blanc, de Jonge, deRijk, & Schaufeli, 2001). This research high-lights the fact that the appropriate unit ofanalysis depends upon the purpose and scopeof the job-design intervention. It is critical

that the unit of analysis be aligned with thespecific needs of the intervention.

Job as the Unit of Analysis

It is possible to reinvent jobs by analyzingthem as a whole. For example, the motiva-tional model (e.g., Hackman & Oldham,1980) has focused on specifying various job

Practitioners should

consider a variety of

approaches to work

design in order to

reduce needed

trade-offs.


8/24


characteristics (e.g., autonomy and feedback)and their impact on motivational levels. Jobsare also a useful unit for designing organiza-tional workflow. The drawback of this level ofanalysis, however, is that it is often too largeand does not provide insight into within-jobvariation. In addition, it provides little direc-tion for specific ways to improve the job(Roberts & Glick, 1981). Using the examplesabove, if the job level of analysis were used incentralizing activities, it would not provideinformation on how jobs could be combinedto achieve efficiencies or on what particular

aspects of a job lead to the in-creased well-being of employees.

Duties as the Unit ofAnalysis

The duties involved in a job canalso be used as the units of analy-sis. Using duties provides more in-formation than the job level andkeeps the number of units man-ageable. With jobs typically con-sisting of 35 duties, however,there are a limited number ofways to combine duties to recon-figure jobs. This may be a goodlevel of analysis for some inter-

ventions, but for other contextsthere may be too few options in terms of re-distributing duties, modifying duties, or com-bining duties into new jobs. This suggeststhat the duty level may still be too gross andlimit the amount of information that can bebrought to bear on the intervention.

Task as the Unit of Analysis

One of the advantages of using the task levelis that a large amount of specific and detailed

information can be obtained about the job.This provides many different options whenredesigning work, because tasks can be re-grouped, eliminated, or changed in a largenumber of ways. For example, tasks fromvarious geographical locations may be com-bined to achieve efficiencies. However, as thenumber of tasks increases, the complexity ofthe analysis increases. For example, if three

jobs consist of 50 tasks each, 1,225 pairedcomparisons would have to be made to un-derstand task interrelationships.

In addition, because task inventories typi-cally consist of as many as 200300 separatetask statements, the possibility for inaccuracyalso increases (Morgeson & Campion, 1997).For example, if an excessively large amount ofinformation is requested of subject-matter ex-perts, it is possible that their motivation toprovide accurate information will be de-creased. Likewise, these high information-pro-cessing requirements could also result in infor-mation overload, which may increase relianceon simplifying heuristics or categorization.

Task Cluster as the Unit of Analysis

Recent research suggests that one way ofovercoming this obstacle is to utilize an in-termediate unit of analysis. As the followingdiscussion suggests, the task level of analysisis often too detailed, and the duty and joblevels of analysis are often too broad. What isneeded in many work-redesign situations is aunit of analysis between a duty and a taskthat allows adequate description of the work,but does not produce information overload.Such an intermediate unit of analysis hasbeen called a task cluster (Cascio, 1995), de-

fined as the smallest collection of logicallyrelated tasks that are normally performed bya single person such that they form a wholeor natural work process (Morgeson & Cam-pion, 2002, p. 593).

A task cluster has four distinguishingcharacteristics. First, a task cluster is broaderthan a task and narrower than a duty, suchthat it typically takes between 10 to 15 taskclusters to adequately describe most jobs. Itprovides a unit of analysis that contains suf-ficient precision to describe a job, but re-

mains manageable in terms of the total num-ber of units. This is particularly importantwhen attempting to understand interdepen-dencies among each pair of units of work.

Second, it is the smallest collection oflogically related tasks normally performed byone person. For example, the job of a data-entry analyst has several tasks that includecreating new electronic database files, saving

One of the

advantages of using

the task level is that

a large amount of

specific and

detailed information

can be obtained

about the job.


9/24


and organizing particular files in specified di-rectories, and creating macros to facilitatedata entry. Although each of these representsa distinct task, they could also be effectivelygrouped into a unit of work called developsdata-entry strategies.

Third, a task cluster is recognized by jobincumbents as a whole or natural piece ofwork. Often, it consists of an entire workprocess or subprocess. Similar to the notion ofidentityin describing an entire job (Hackman& Oldham, 1976), the task cluster is seen asdistinct from other units of work and as com-plete in itself, thereby constituting a naturalsegment of work. This allows jobs to be rein-vented using units high in identity, therebyfacilitating motivation and satisfaction.

Fourth, task clusters are composed of tasks

that have at least moderate levels of task inter-dependence (Wong & Campion, 1991). Thereare many advantages to grouping interde-pendent tasks. For example, the interdepend-ence of tasks increases the likelihood that per-formance on one task will provide some formof feedback about the performance on a priortask. In addition, when interdependent tasksare performed by one individual, it requiresfewer resources to coordinate the tasks than ifmultiple people were responsible for complet-ing them. Using our running example, a task

cluster may be able to provide the right level ofanalysis to combine data-entry strategies fromeach geographical location.

Thus, overcoming this obstacle requirespractitioners to use the unit of analysis thatis best suited to the scope of the interven-tion. The task cluster concept provides onepossible solution. Practitioners should fur-ther evaluate the value of task clusters for re-design purposes across a range of occupa-tions and interventions. To aid practitioners,future research should seek to provide advice

for choosing among the units, and discussimplications for using specific units on thework-design results obtained.

Obstacle #4: It Is Difficult to Predictthe Nature of a Job Before It Exists

Work redesign involves the creation of newjobs or modification of existing ones. In

either case, one objective is to predict thecharacteristics of the resulting job prior to itsactual redesign. This objective is an obstaclefor practitioners because our capacity formaking accurate predictions is inadequate.More important, we must ensure that em-ployees perceive the new job as we intended,something that does not always occur. Onereason why conceptual job redesign does notalways match actual redesign is that the latteroften is based on changes in the jobs com-ponent parts. Our inability to predict arises,at least partially, from the fact that the wholejob generally is not equal to thesum of the parts and that certaincombinations of changes mayproduce unforeseen outcomes.

Several factors help explain

why knowledge of individualcomponents may prove insuffi-cient for predicting the new jobas a whole. First, some job-levelcharacteristics simply do not existat lower levels. For example, taskvariety only exists when tasks areviewed together. An evaluation ofindividual tasks would be insuffi-cient for understanding howcombinations of tasks impact the variety ofthe entire job. Similarly, the degree to which

an individual completes a whole portionof work (i.e., identity) cannot be judged bylooking at single parts of the job in isolation.

Second, looking only at individual partsof jobs ignores the interdependencies amongthem. That is, tasks are interdependent whenthe inputs, processes, or outputs of one taskaffect or depend on the inputs, processes, oroutputs of other tasks within the same job(Wong & Campion, 1991). Because tasks maybe interdependent, the evaluation of thetasks taken as a group can be quite different

from their evaluation in isolation. For exam-ple, the level of feedback present in a jobmay depend on whether an individual per-forms several steps of a process. The presenceof feedback may only be evident when oneconsiders interdependencies among theparts. Likewise, the interdependenciesamong jobs and work systems within the or-ganization impact the differentiation in HR

We must ensure

that employees

perceive the new

job as we intended,

something that does

not always occur.


10/24


strategy that can be effectively employedwithin an organization (Delery & Shaw,2001).

In support of these notions, Campionand Stevens (1991) and Wong and Campion(1991) found that the motivational value ofa job could be only modestly predicted bysumming the motivational value of the indi-vidual tasks. For example, Campion andStevens (1991) had experts rate a set of 40tasks on dimensions reflecting the extent towhich the tasks represented motivational,mechanistic, biological, or perceptual job-de-sign models. Subjects were then asked togroup the 40 tasks into jobs, and independ-ent subjects rated each of thejobs in terms ofthe four job-design models. They foundsmall relationships between the job and the

task ratings for three of the four job-designmodels (.10 to .22) in this initial study, andfor only two of the four in a follow-up. Thisfinding suggests that simply having knowl-edge of the tasks does not allow one to fullypredict the job as a whole.

Wong and Campion (1991) found that ameasure of task interdependence could en-hance the prediction of the motivationalvalue of the job beyond the sum of the mo-

tivational value of the tasks. In a sample of188 employees in 67 jobs, employees ratedtheir jobs in terms of the motivational job-design model, and independent expertsrated the tasks on their motivating value, in-terdependence, and similarity. The resultsshowed that the motivational design of thetasks was only modestly related to the moti-vational design of the jobs, and task similar-ity added little to the prediction. However,prediction was improved by considering in-terdependencies among the tasks. Specifi-cally, as interdependencies among tasks in-creased, satisfaction with the job alsoincreased, but only up to a point (see Figure1). Very high levels of interdependence wereassociated with lower ratings of the motiva-tional design of the jobs. It may be that ex-

treme levels of interdependence result in ei-ther role overload or narrowly designed jobswith limited stimulation (Wong & Campion,1991).

Similarly, for a new job, practitionersmay find it difficult to predict the conse-quences that occur from making changes incombination. For example, consider the im-plementation of manufacturing technologiesdesigned to increase attentional demand and

FIGURE 1. Relationship between motivational job-design and task level measures (from Wong & Campion,1991).


11/24


cost responsibility for operators. Research in-dicates that the combination of high levelsof attentional demand coupled with addedcost responsibility causes psychologicalstrain compared to employees working inother combinations of the variables (Martin& Wall, 1989). Practitioners may find it diffi-cult to predict the combinations that willproduce the undesirable outcomes prior tothe implementation of new technologies.

Research into the role of interdepend-ence and combinations of changes hasadded to our understanding of both work-design theory and practice. More important,this research helps practitioners overcomethe obstacle of predicting the nature of jobs.A partial solution to the problem of predict-ing the nature of a job before it exists is to

first take into account both the motivationalvalue and the interdependencies of tasks.Once the job is designed, it must be viewedin its entirety to fully understand the natureof the job being created, and the likely reac-tion of employees.

Practitioners should consider a widerange of possible interdependencies betweenjobs, such as the 14 types of interdependencedefined by Wong and Campion (1991), in-cluding inputs (e.g., materials and informa-tion), processes (e.g., scheduling and super-

vision), and outputs (e.g., goals andperformance). This investigation would alsohelp practitioners to align the job designwith the organizations HR architecture. Forexample, increased interdependenciesamong jobs would be consistent with thecollaborative HR configuration proposed byLepak and Snell (1999). Likewise, practition-ers should consider the different combina-tions of changes and the potential impact ofeach change on employees. Accounting forthe numerous combinations may help mini-

mize the undesirable outcomes of redesignedjobs.

Obstacle #5: Individual DifferencesComplicate Job Redesign

Individuals differ in terms of the attitudesand beliefs they hold, what they value, andhow they respond to their environment. To

the extent people also differ in how they re-spond to characteristics of their work envi-ronment, an obstacle to designing work ef-fectively is created. This obstacle becomesparticularly relevant when incumbents areinvolved in the work design. Their input intothe design may be based on their own indi-vidual preferences, which may not holdacross employees. Individual differences alsocomplicate job redesign when organizationsare redesigning into teams, as individual dif-ferences affect a variety of team outcomes.Fortunately, several studies have used themotivational model to studyteams, thus providing insight topractitioners on how individualdifferences may influence re-sponses to work design.

Early Moderator Research

The early work on moderatorsoriginated with Turner andLawrence (1965), who found evi-dence that urban/rural back-ground moderated the relation-ship between job characteristicsand satisfaction. Individuals fromrural backgrounds responded more posi-tively to enriched work. At about the same

time, other researchers (Blood & Hulin,1967; Hulin & Blood, 1968) investigatedalienation from middle class norms andfound limited evidence for the moderatoramong blue-collar respondents. Others alsofound significant moderating effects for jobinvolvement (Ruh, White, & Wood, 1975)and need for achievement (Steers, 1975).However, still other researchers investigatingmoderators such as community size (Shep-ard, 1970) and Protestant Work Ethic (Stone,1975, 1976) found little to no evidence

(White, 1978).

Growth Need Strength

The most commonly studied moderator ofthe work designwork outcome relationshipis Growth Need Strength (GNS). GNS is thepreference or need individuals have for stim-ulating and challenging work. The basic

Individuals differ in

terms of the attitudes

and beliefs they hold,

what they value, and

how they respond to

their environment.


12/24


premise is that motivation and satisfactionresult from a goodness of fit between thetask characteristics and the needs of the em-ployees, while the relationship between mo-tivating job design and job satisfaction isstrongest for high-GNS individuals (Hack-man & Oldham, 1980). There has been con-siderable research investigating this premise,with some studies finding evidence for GNSas a moderator (Hackman & Lawler, 1971;Hackman, Pearce, & Wolfe, 1978) and othersfinding a weaker relationship (Lawler, Hack-man, & Kaufman, 1973).

Several meta-analytic studieshave summarized this researchand have reached optimistic con-clusions about the moderatingrole of GNS. For example, after

conducting a meta-analysis of 28studies, Loher, Noe, Moeller, andFitzgerald (1985) concluded thatGNS was useful as a moderatingvariable of the job designjob sat-isfaction relationship. Fried andFerris (1987) also suggested thatGNS moderated the relationshipbetween motivational job designand job performance, althoughthey only found five studies thatactually examined this relation-

ship. The meta-analytic evidence,however, should be balanced against the over-all criticisms leveled against this research dueto its overreliance on self-report measures(Johns, Xie, & Fang, 1992; Roberts & Glick,1981), consistency artifacts (Podsakoff &Organ, 1986), and potential for response-re-sponse bias (Salancik, Staw, & Pondy, 1980).

Other Moderators

Campion (1988) investigated whether pref-

erences for work designed from each of fourdifferent models would moderate responsesto jobs designed from those models, butfound only limited support for preferences asa moderator. It is also possible that em-ployee-ability levels influence reactions tojob-redesign efforts. If the cognitive abilityrequired for the job is beyond that possessedby the individual, change may be perceived

less positively. For example, Schneider, Re-ichers, and Mitchell (1982) and Dunham(1977) found significant relationships be-tween motivational characteristics of jobsand various ability requirements. From themultidisciplinary perspective, Campion(1989) found that motivational job designhas a positive relationship with a wide rangeof mental ability requirements and that jobsdesigned from a mechanistic or a perceptualperspective were negatively related to mentalability requirements. Although it remains animportant research question, there is adearth of research specifically investigatingthe moderating role of employee abilities(Fried & Ferris, 1987).

Other researchers have hypothesized thatthe quality of interpersonal relationships at

work may moderate the impact of job designon job attitudes, arguing that when workersenjoy satisfying relationships on the job itminimizes the detrimental impact of nega-tive job design. In one particular study (Fretz& Leong, 1982), for example, the results weregenerally in the predicted direction, butmost relationships were not significant. Inaddition, Oldham (1976) studied the moder-ating role of supervisory and coworker satis-faction on the relationship between job de-sign and intrinsic motivation. Other studies

have found mixed (Abdel-Halim, 1979;Johns et al., 1992; Oldham, Hackman, &Pearce, 1976) or negative results, leaving therole of interpersonal context as a moderatorin question.

The weight of the evidence suggests thatindividual differences may influence the wayindividuals respond to their work. These dif-ferences, however, do not appear to presentan obstacle to job-design interventions fortwo reasons. First, in most instances wherejobs are being designed for multiple employ-

ees, it is best to design jobs in accordancewith the average or typical employee. If jobsare tailored to the individual preferences ofeach current incumbent, the jobs may not besuited to future incumbents who might pos-sess different preferences. Furthermore, re-designing the job for each new employee isimpractical, and predicting the preferencesof future employees is likely to become more

In most instances

where jobs are

being designed for

multiple employees,

it is best to design

jobs in accordance

with the average or

typical employee.


13/24


difficult with changes in labor market demo-graphics. In addition, if a company decidesto move to a team-based work design, it isoften impossible to exclude certain employ-ees from working in teams, and thus workcannot be redesigned around individual jobholders. However, practitioners can take intoaccount the members individual differenceswhen redesigning work as well as their pref-erences for working in a team (Hollenbeck,DeRue, & Guzzo, 2004).

Second, the relationships between thejob-design models and their outcomes tendto be in the same direction for all employees,even if they differ in strength between em-ployees. For example, although some em-ployees may respond more positively to themotivational model than others, the rela-

tionship is rarely negative. That is, typicallyall employees respond positively to motivat-ing work, but some employees respond morepositively than others (White, 1978). Re-search on GNS is a good illustration. Eventhose employees low in GNS showed smallincreases in job satisfaction in response tothe motivating job characteristics (Loher etal., 1985). In addition, there is evidence thatpeople generally prefer work that is designedto be motivating. Campion and McClelland(1991) found that individuals generally pre-

ferred jobs designed from the motivationalperspective and not the perceptual perspec-tive, and were ambivalent about jobs de-signed from the mechanistic or biologicalperspective.

Understanding individual differences iskey for practitioners when redesigning workeither for individuals or for teams. When de-signing jobs for individual workers, practi-tioners should account for individual differ-ences. However, when designing jobs formultiple employees, the role of individual

differences should be considered, but notviewed as a major obstacle. In these in-stances, practitioners should examine thepreferences of large groups of employees(e.g., in a specific occupation) rather thanthe preferences of particular individuals.When redesigning work into teams, practi-tioners may be able to align the individualdifferences of team members with the

desired outcomes of the team. In addition,practitioners can utilize selection, training,and structures such as rewards to enhanceteam functioning (Hollenbeck et al., 2004).

Obstacle #6: Job Enlargement Can

Occur without Job EnrichmentOrganizations often redesign jobs to en-hance their motivational value (Herzberg,1966). However, this accomplishment is notalways an easy matter. Sometimes jobs areenlarged but not enriched, and the motiva-tional benefits do not occur. Thislimitation creates an obstacle forpractitioners because althoughthe distinction between job en-largement and enrichment may

be fairly straightforward, it maybe difficult to predict if employ-ees will perceive the changes asenrichment or as enlargement.For example, research on down-sizing has demonstrated thatmanagement perceives changessuch as increased responsibilityand decision making as job en-richment, while employees per-ceive the same changes as job en-largement and increased role overload

(Tombaugh & White, 1990). In addition,practitioners may come to the false-negativeconclusion that job enrichment did notbring the intended benefits, when in realitythe lack of improvement was due to the in-adequate implementation of the enrichmentprinciples.

Job enlargementrefers to increasing the mo-tivational value of a job through the perform-ance of a greater number and variety of simi-lar-level tasks (Lawler, 1969). Herzberg (1968)referred to this combination as horizontal load-

ing, and Lawler (1969) labeled it horizontal en-largement. Horizontal enlargement has beencriticized, however, because it may do littlemore than replace one boring task or responsi-bility with two or three boring tasks or re-sponsibilities. Job-enrichment strategies at-tempt to overcome these weaknesses.

Job enrichmentrepresents the vertical load-ing of jobs (Herzberg, 1968), such as the

Understanding

individual

differences is key for

practitioners when

redesigning work

either for individuals

or for teams.


14/24


adding of tasks for planning and controllingwork (Lawler, 1969). It involves adding tasksand decisions that typically are performed bysomeone higher up in the organizational hi-erarchy (e.g., a supervisor). Hackman and Old-ham (1975) more fully articulated the notionof job enrichment. Their model delineates thepsychological states (e.g., experienced mean-ingfulness of work, felt responsibility) thatmediate the relationship between the job-de-sign principles (e.g., autonomy, skill variety)and such outcomes as intrinsic motivation,job satisfaction, and performance. Recent re-

search has also investigated therole of job enrichment on out-comes such as role breadth self-ef-ficacy, or an employees perceivedcapability of carrying out a

broader and more proactive role.Results indicate that increased jobenrichment leads to the develop-ment of greater self-efficacy (Axtell& Parker, 2003; Parker, 1998).

One of the obstacles facedduring a work-design interventionis how to actually enrich a job in-stead of simply enlarging it. Re-cent research has made progressin understanding job enrichmentversus job enlargement. Two stud-

ies on role breadth self-efficacyfound that the key to increasing

an employees self-efficacy was the degree ofenrichment in the work redesign rather thanthe extent of enlargement (Axtell & Parker,2003; Parker, 1998). In fact, one of the stud-ies found that after controlling for greater in-volvement, job enlargement actually had anegative effect on role breadth self-efficacy.The implications support the idea that simplyexpanding the breadth of tasks (enlarging ajob), without increasing motivational aspects

(enriching the job), might actually reduce anemployees role breadth self-efficacy. Asnoted by the authors, [T]his finding is im-portant because organizations often believethey are empowering the workforce when infact all they are doing is enlarging their tasks(Axtell & Parker, 2003, p. 126).

One recent form of enrichment, knowl-edge enrichment, may aid practitioners in

job-design interventions. Knowledge enrich-ment involves adding requirements for un-derstanding job-related procedures or rules.As a research construct, it appears relatedboth to job satisfaction and performance.Campion and McClelland (1993) collecteddata on a wide range of costs and benefitsfrom over 400 employees whose jobs hadbeen redesigned at a financial service institu-tion. The longitudinal data indicated thattask enlargement resulted primarily in costs,whereas knowledge enrichment resulted pri-marily in benefits.

An additional important distinction forpractitioners to consider is the implementa-tion of the job-enrichment changes. Theprocess used to implement the change,whether top-down or bottom-up, may im-

pact whether employees perceive the changeas enrichment or enlargement. A top-downapproach to enrichment (i.e., the job modifi-cations are prescribed by management toemployees) may differ substantially in its ef-fects from a bottom-up approach (i.e., jobmodifications are suggested by employees tomanagement). One might surmise that iden-tical changes made to enrich jobs may havedifferent effects depending on whether ornot they were created and implemented in atop-down or bottom-up manner, with the

bottom-up approach likely resulting ingreater employee acceptance.

Another question faced by practitionersis whether there is a limit to the amount thata job can be enriched. It is likely that there isa point at which adding more tasks, respon-sibilities, or knowledge requirements leads todecrements in performance and satisfaction.This may occur if the changes in the job re-quire knowledge, skills, and abilities (KSAs)that the incumbents do not possess or if thechanges overload the incumbent with too

many tasks to perform effectively. We referto such a situation asjob engorgement. An ex-ample is when the addition of tasks shouldhave a positive effect on satisfaction and mo-tivation, but instead increases role ambigu-ity, role overload, and work stress (e.g.,Abdel-Halim, 1978; Jamal, 1984). Researchindicates that increases in role ambiguity,role overload, and role conflict result in un-

It is likely that there

is a point at which

adding more tasks,

responsibilities, or

knowledge

requirements leads

to decrements in

performance and

satisfaction.


15/24


desired outcomes such as decreased satisfac-tion and increased intent to leave(Tombaugh & White, 1990). These situationsmay be especially common for new employ-ees or high-potential employees who mayfind it difficult to refuse additional tasks andresponsibilities.

Thus, in attempting any work-redesignintervention, it is necessary for practitionersto explicitly consider how to enrich the jobrather than simply enlarge it. One suggestionfor practitioners is to implement work de-signs that involve decision-making influenceand control rather than simply increasingthe number of tasks. This may lead to desiredoutcomes such as employees carrying outmore proactive, integrative, and interper-sonal tasks, which are important in todays

organizations (Axtell & Parker, 2003). Addi-tional research and greater judgment on thepart of practitioners is needed, however, tofurther explicate the empirical distinctionsbetween job enlargement, job enrichment,and knowledge enrichment, and how toavoid job engorgement in the process. Thisresearch will further aid practitioners in de-signing jobs to enhance their motivationalvalue.

Obstacle #7: New Jobs Need to Be

Created as Part of Growth orDownsizing

One of the defining characteristics of themodern workplace is rapid change. For prac-titioners, the obstacle created by rapidchange is how jobs should change in re-sponse to growth or contraction. An essen-tial feature of this obstacle is the large-scalenature of the change, leading to increases ordecreases in the number of employees avail-able to perform the work. In addition, prac-

titioners also must consider the conse-quences of contraction on employees whoremain in the organization. As mentionedpreviously, research indicates that manage-ment may view downsizing as increasing re-sponsibility and decision making, while em-ployees report increased stress due to roleoverload, role conflict, and role ambiguity.The result of this perceptual difference is

higher levels of employee dissatisfactionalong with increased intent to leave the or-ganization (Tombaugh & White, 1990). For-tunately for practitioners, numerous optionsare available on how to redesign work in themidst of growth or contraction.

Growth

As organizations grow, so do the numberand type of jobs that are used to organizethe work. What are the ways in which an or-ganization can create new jobs to keep upwith the growth, and what arethe likely consequences for theemployees? There are a numberof alternatives in new job growth(Figure 2), and predictions re-

garding outcomes can be madebased on interdisciplinary job-design theory.

Uncoupling involves separat-ing tasks into two jobs at thesame level. It is the type of ex-pansion that is most likely underconditions of gradual growth, orwhen the additional work to bedone is not qualitatively differentfrom current job demands. For example, as aretail outlet prepares for seasonal fluctua-

tions in the number of customers and vol-ume sold, it must create new jobs. It mayhire additional sales associates whose dutiesdo not differ substantially from other salesassociates. In this way, the volume of work issimply divided between the initial jobs andthe new jobs. Since the jobs are at the samelevel, outcomes such as satisfaction, ability,training, and compensation requirementsare likely to remain the same.

Unstackinginvolves separating tasks intotwo jobs at different levels. The creation of a

manager position is an example of unstack-ing. This approach is most likely used ifgrowth has created coordination or controlproblems within the workforce. Complexduties (such as planning, monitoring, and re-porting) may be assigned to a manager, withbasic production or service responsibilitiesgiven to lower-level employees. Practitionersshould note that this approach to job cre-

For practitioners,

the obstacle

created by rapid

change is how jobs

should change in

response to growth

or contraction.


16/24


ation may result in lower-level employeeshaving relatively lower job satisfaction,training, and compensation requirementscompared to upper-level employees. On theother hand, a potential benefit for the or-ganization is that the lower-level employeesmay be more efficient due to opportunitiesfor specialization and repetition.

Segmenting results from taking portionsout of other jobs and combining them into anew job. The organizational level of the newjob may be higher or lower. For example,when small companies grow, it is often thecase that the HR tasks performed by a varietyof employees are pulled together to form adedicated HR job. Thus, instead of each em-

FIGURE 2. Alternatives in new job growth.


17/24


ployee being responsible for some recruiting,a single job would be created to handle allthe recruiting responsibilities. These newjobs should create efficiencies, because tasksare grouped by specialty. The influence onmotivation is uncertain, however. These jobsmay motivate employees due to the interde-pendence among the tasks and resultinghigh task identity, or these jobs may fail tomotivate due to excessive repetition and thepotential for overspecialization.

Emergentjobs result from the merging oftasks that had not been previously per-formed or that had been contracted out. Asorganizations respond to environmental de-mands, new tasks may be necessary to meetcustomer expectations, improve businessprocesses, or use modern technology. In ad-

dition, organizations may perform functionsinternally that had previously been out-sourced. Although these jobs may lack effi-ciency in the short term as roles, procedures,and best practices are established, they mayyield benefits in the long term in terms ofcustomized service and quicker responsetime. Allowing job incumbents a degree ofautonomy in defining these unique roles isalso likely to yield motivation and satisfac-tion benefits.

Finally, in circumstances of opportunistic

hiring, jobs are created to match the skills ofnew organizational members. In an effort torecruit and hire high-potential employees,jobs may be created that are tailored to theskills and preferences of individual employ-ees. These jobs have clear developmentalbenefits, because new hires are given a de-gree of latitude and individual discretion indefining the job. This type of hiring maybring motivational benefits as well as in-creased role innovation. Costs may also beincurred, however, because the job may not

be well integrated with other jobs, and someindividuals may flounder with an unstruc-tured job.

Contraction

Another obstacle organizations face is the re-design of jobs due to restructuring or down-sizing. There are many possible conse-

quences of this process in terms of the designof jobs and impact on employees. Better job-design decisions may be possible by knowingthe possible alternatives and evaluating thepotential outcomes based on an interdisci-plinary job-design perspective (Figure 3). Inaddition, if noncore jobs are identified, theyare prime candidates for contraction or out-sourcing.

Jobs that remain following downsizingmay be expanded to perform tasks that hadpreviously been part of the jobsthat were eliminated. It is likelythat some jobs may be enlargedthrough the addition of tasks thatare on a similar level in terms oftheir responsibilities. At best, thelikely result of such enlargement

is jobs that have a little more va-riety. However, as noted previ-ously, enlargement without en-richment often is not motivating.

Similarly, some jobs may beoverloadedif the employees are ex-pected to accomplish much of thework that was previously per-formed by other jobs. The likelyresults are dissatisfaction,turnover, and reduced performance. How-ever, there are possible positive conse-

quences of both enlargement and overload-ing. In addition to labor cost savings, worksystems may be streamlined, and unneces-sary tasks may become more visible anddropped.

The remaining jobs also may be enrichedthrough the addition of higher-level tasks ortasks that require additional ability andtraining. A typical example occurs whenmanagement positions are eliminated andtheir responsibilities are given to workteams. This type of change can have a posi-

tive effect on motivation and satisfaction, al-though costs may be incurred due to in-creased errors and the need for additionaltraining, along with efficiency losses, whilethe team assimilates the new responsibilities.

As organizations are restructured andjobs are eliminated, it is likely that certaintasks simply may not be performed becausethe worker is no longer there. Also, some

Jobs that remain

following

downsizing may be

expanded to

perform tasks that

had previously been

part of the jobs that

were eliminated.


18/24


jobs may be impoverished as a result of thedownsizing. Alternatively, some individualswill have their tasks or duties reduced. Inparticular, managers may suffer a reducedspan of control due to employee layoffs,moving them to other areas, or cutting proj-

ects. The result of impoverishment can be areduction in the motivational value of thejob due to the performance of lower-leveltasks or a smaller job scope.

Finally, still other jobs may be reconfig-ured. The reconfiguration of jobs results from

FIGURE 3. Alternatives in job contraction.


19/24


the shifting of tasks, duties, or responsibilitiesamong jobs. In an effort to redistribute work-load, avoid impoverishment, and preventoverloading, organizations may reconfigurejobs to more evenly distribute the amount ofwork among the employees and jobs that re-main after the downsizing effort. The conse-quences of reconfiguration for job design willdepend on the precise nature of the changes.This is a prime opportunity to reconfigurejobs to make them more closely related to theorganizations strategic goals and to furtherits human resource management strategy.

Another challenge resulting from con-traction that practitioners face is the conse-quences of the change to those who remainin the organization. Research on downsizingprovides practitioners with evidence on how

to best manage the change for the survivorsin the organization. Parker, Chmiel, andWall (1997) investigated the long-term im-plications of strategic downsizing for em-ployee well-being as a function of its effecton work characteristics. Results providestrong support that increased clarity aboutroles and responsibilities and increased par-ticipation in the change are particularly im-portant predictors of both strain and job sat-isfaction. The study also indicates thatincreased control is positively associated

with job satisfaction (Parker et al., 1997).This research provides several guidelines

for practitioners redesigning jobs duringcontraction. First, to offset the negative con-sequences of demand, practitioners shouldestablish clear roles and responsibilities foremployees and inform and involve employ-ees in the change. Second, in order to en-hance employee-well being, practitionersshould enhance levels of control over thetiming and methods of work. Finally, practi-tioners should pay attention to the design of

work and the wider organizational contextin order to enhance an organizations abilityto achieve contraction without incurringlong-term negative consequences for em-ployees (Parker et al., 1997).

Whether in the midst of growth or con-traction, several options are available topractitioners for redesigning work. The keywill be for practitioners to align redesign

effort with the strategic goals of the organi-zation and its human resource managementstrategy. In addition, practitioners can utilizeinterdisciplinary job-design theory to con-sider the consequences of each alternative.

Obstacle #8: Long-Term Effects MayDiffer from Short-Term Effects

Although seldom investigated, the long-termeffects of work-design interventions may notbe the same as the short-term ef-fects. It is possible that the short-term evaluation of an interven-tion may indicate that desiredchanges in satisfaction or produc-tivity have been achieved, butthat these positive effects dissipate

with time. Similarly, there may beunanticipated long-term costs orbenefits associated with the inter-ventions that were not present atthe short-term evaluation. Thus,practitioners face the obstacle ofhow to sustain the benefits fromjob design over the long term.Practitioners may need to measurechanges in jobs and evaluate thesechanges on a periodic basis. Overcoming thisobstacle is critical to fully understanding the

strategic implications of changes in HR man-agement systems (Delery & Shaw, 2001).

As the majority of field research evaluat-ing the impact of job-design interventions isconducted over a period of six months to ayear, studies that evaluate these interven-tions over the long term are informative. Forexample, Campion and McClelland (1991,1993) conducted both a short-term evalua-tion and a two-year follow-up to a job-re-design intervention. They found that al-though jobs designed from the motivational

model were more satisfying in the shortterm, certain applications of the motiva-tional model were negatively related to satis-faction over a longer period. Other researchdemonstrates that potentially negative out-comes may not appear immediately after theintervention. For example, Parker (2003)evaluated the longitudinal effects of leanproduction and found that organizational

Although seldom

investigated, the

long-term effects of

work-design

interventions may

not be the same as

the short-term

effects.


20/24


commitment did not differ between groupsat the outset, but after three years, one grouphad lower commitment than the othergroups combined.

Short-term and long-term effects maydiffer for several reasons. First, the efficiencyand quality of work may increase over timeas employees gain experience with the task,gain a better understanding of performancestandards, increase proficiency at catchingerrors, and become more aware of the mostrelevant sources of feedback. Second, train-ing costs decrease over time. Immediately

following changes in job design,training on the new tasks, respon-sibilities, and technologies is nec-essary for employees and supervi-sors (Hackman & Oldham, 1980).

With time, however, these costsdecrease as employees completethe training (Campion & McClel-land, 1993).

Third, the novelty of changemay be short-lived. That is, thechanges themselves may initiallyappeal to employees who enjoythe increased attention, thechange of pace, and the new tasks

that result in increased interest in and satis-faction with the job. Over time, however,

employees may become habituated to themodified job, returning job attitudes to theirbaseline levels (Homans, 1950; Roethlis-berger & Dickson, 1939). Thus, boredomwith the work may return in the long run.Fourth, too much enrichment may decreaseemployee preferences for enriched work inthe future. Campion and McClelland (1993)illustrated this when they found that a de-creased preference for enriched work wasseen after the enrichment intervention.

Fifth, as they habituate to the new work

environment, employees may realize that thenew jobs are more difficult and require moreskill. Demand for higher compensation islikely to follow. This occurs as employees re-alize they are doing more highly skilled work.Their value in the labor market is increasedbecause of the new skills, and replacement ofemployees forces organizations to pay marketwages (Campion & Berger, 1990).

Sixth, individuals may act to simplify ajob that has been enlarged. For example,Campion and Stevens (1991) found that indesigning jobs, people tend to group similartasks in such a way that reflects the mecha-nistic perspective. That is, in an attempt tominimize work, people tend to break workdown into simple repetitive tasks. There isalso anecdotal evidence that this occurred inactual jobs (Campion & McClelland, 1993).

Finally, the form of the intervention it-self may evolve over time. In Campion andMcClelland (1993), the initial task enlarge-ment evolved to either task enlargement orknowledge enrichment at different parts ofthe company. As noted previously, task en-largement had negative effects in the longterm, whereas knowledge enrichment had

primarily positive effects.All this evidence suggests that long-termoutcomes are likely to diverge from short-term outcomes. For practitioners, this createsan obstacle in redesigning work, since the out-comes may dissipate and costs may increaseover time. Thus, it is critical that practitionersconduct more extended evaluations of jobchanges. Furthermore, from the planningstages of the job-design intervention, practi-tioners should consider the long-term effectsof job design as well as the short-term effects.

Conclusion

Work-design issues are ubiquitous in organi-zations, and practitioners continue to faceseveral challenges, or obstacles, when imple-menting work-design changes. We set out inthis article to examine eight obstacles thatmay impede practitioner efforts to redesignwork. In the discussion, we reviewed work-de-sign research and provided direction to aidpractitioners in dealing with the obstacles. Al-

though we have not completely overcomeeach obstacle, we hope the discussion helps toclarify the issues for practitioners and pro-vides guidance on what pitfalls to avoid for asuccessful job-design implementation.

Going forward, we feel continued dia-logue among HR management professionalsand researchers could provide significant in-sight on strategies to surmount these obsta-

It is critical that

practitioners

conduct more

extended

evaluations of job

changes.


21/24


cles and implement even more successfulwork-design interventions. Continued re-search on work design is particularly criticalbecause it influences a large number of im-portant outcomes (e.g., satisfaction, effi-ciency, performance) and accounts for alarge amount of statistical variance in theseoutcomes. Work-design research represents atremendous opportunity to enhance thecontribution of HR research to organiza-tions. In addition, encouraging the theory-

based redesign of work from an interdiscipli-nary perspective will help us overcome theseobstacles and meet the challenges of design-ing work in the twenty-first century.

Note

This paper is based on the Presidential Addressthe first author delivered at the 1996 AnnualMeeting of the Society for Industrial and Organi-zational Psychology, San Diego, California.

MICHAEL A. CAMPION is a professor of management at Purdue University (for the last

19 years). Previous industrial experience includes eight years at IBM and Weyerhaeuser

Company. He has an MS and PhD in industrial and organizational psychology. He has

over 85 articles in scientific and professional journals and has given over 170 presenta-

tions at professional meetings, on such topics as interviewing, testing, job analysis, work

design, teams, training, turnover, promotion, and motivation. He is among the ten mostpublished authors in the top journals in his field for both the last two decades. He is past

editor of Personnel Psychology(a scientific research journal) and past president of the

Society for Industrial and Organizational Psychology.

Dr. TROY V. MUMFORD received his PhD in organizational behavior and human resource

management from the Krannert Graduate School of Management at Purdue University.

He currently teaches courses in human resource management, including compensation,

performance management, and organizational behavior. His current research interests

include determinants of team member effectiveness, employee choice of flexible bene-

fit options, and multilevel leadership skill requirements. Dr. Mumford has also been in-

volved in a variety of independent consulting and applied research projects for a num-

ber of organizations in the biomedical, manufacturing, financial, and airline industries as

well as the U.S. government. These projects span many HR functions, including job

analysis, selection, recruiting, training, compensation, HR planning, process analysis,

and performance appraisal.

FREDERICK P. MORGESON is currently an associate professor of management at the Eli

Broad College of Business at Michigan State University. He received his PhD in industrial

and organizational psychology from Purdue University. Dr. Morgeson teaches and does re-

search in human resource management and organizational behavior. His research involves

attempting to understand the nature of work, which includes examining how jobs and

work are measured, as well as how work is designed, including newer forms of work de-

sign such as teams. In addition, Dr. Morgeson studies the effectiveness of different staffing

techniques and the role leadership plays in high-risk and team-based environments. This

research has been published in the Academy of Management Review, American Psychol-ogist, the Consulting Psychology Journal: Practice and Research, the Journal of Applied

Psychology, the Journal of Business and Psychology, the Journal of Organizational Be-

havior, the Journal of Quality Management, and Personnel Psychology. In addition, he has

authored or coauthored several book chapters, presented over 50 papers at professional

meetings, is a member of the Personnel Psychologyand Journal of Management editorial

boards, and is a recipient of the 2005 American Psychological Association Distinguished

Scientific Award for Early Career Contribution to Psychology in Applied Psychology.


22/24


REFERENCES

Abdel-Halim, A. A. (1978). Employee affective re-

sponses to organizational stress: Moderating ef-

fects of job characteristics. Personnel Psychology,

31, 561579.

Abdel-Halim, A. A. (1979). Individual and interper-

sonal moderators of employee reactions to job

characteristics: A reexamination. Personnel Psy-

chology, 32, 121137.

Arthur, J. B. (1992). The link between business strat-

egy and industrial relations systems in American

steel minimills. Industrial and Labor Relations Re-

view, 45, 488506.

Arthur, J. B. (1994). Effects of human resource sys-

tems on manufacturing performance and turnover.

Academy of Management Journal, 37, 670687.

Astrand, P. O., & Rodahl, K. (1977). Textbook of work

physiology: Physiological bases of exercise (2nd

ed.). New York: McGraw-Hill.

Axtell, C. M., & Parker, S. K. (2003). Promoting role

breadth self-efficacy through involvement, work re-

design and training. Human Relations, 56, 113131.

Barnes, R. M. (1980). Motion and time study: Design

and measurement of work (7th ed.). New York: Wiley.

Blood, M. R., & Hulin, C. L. (1967). Alienation, environ-

mental characteristics, and worker responses.

Journal of Applied Psychology, 51, 284290.

Campion, M. A. (1988). Interdisciplinary approaches to

job design: A constructive replication with exten-

sions. Journal of Applied Psychology, 73, 467481.

Campion, M. A. (1989). Ability requirement implica-

tions of job design: An interdisciplinary perspec-tive. Personnel Psychology, 42, 124.

Campion, M. A., & Berger, C. J. (1990). Conceptual in-

tegration and empirical test of job design and

compensation relationships. Personnel Psychol-

ogy, 43, 525553.

Campion, M. A., & McClelland, C. L. (1991). Interdisci-

plinary examination of the costs and benefits of

enlarged jobs: A job design quasi-experiment.


Campion, M. A., & McClelland, C. L. (1993). Follow-up

and extension of the interdisciplinary costs and

benefits of enlarged jobs. Journal of Applied Psy-

chology, 78, 339351.

Campion, M. A., Medsker, G. J., & Higgs, A. C. (1993).

Relations between work group characteristics and

effectiveness: Implications for designing effective

work groups. Personnel Psychology, 46, 823850.

Campion, M. A., & Stevens, M. J. (1991). Neglected

questions in job design: How people design jobs,task-job predictability, and influence of training.

Journal of Business & Psychology, 6, 169191.

Campion, M. A., & Thayer, P. W. (1985). Development

and field evaluation of an interdisciplinary meas-

ure of job design. Journal of Applied Psychology,

70, 2943.

Cascio, W. F. (1995). Whither industrial and organiza-

tional psychology in a changing world of work?

American Psychologist, 50, 928939.

Cummings, T. G. (1978). Self-regulating work groups:

A socio-technical synthesis. Academy of Manage-

ment Review, 3, 625634.

Delery, J. E., & Shaw, J. D. (2001). The strategic man-

agement of people in work organizations: Review,

synthesis, and extension. Research in Personnel

and Human Resources Management, 20, 165197.

Deming, W. E. (1986). Out of the crisis. Cambridge,

MA: Massachusetts Institute of Technology, Center

for Advanced Engineering Study.

DeVor, R. E., Chang, T. H., & Sutherland, J. W. (1992).

Statistical quality design and control. Upper Sad-

dle River, NJ: Prentice-Hall.

Dunham, R. B. (1977). Relationships of perceived jobdesign characteristics to job ability requirements

and job value. Journal of Applied Psychology, 62,

760763.

Edwards, J. R., Scully, J. A., & Brtek, M. D. (1999). The

measurement of work: Hierarchical representation

of the multimethod job design questionnaire. Per-

sonnel Psychology, 52, 305334.

Edwards, J. R., Scully, J. A., & Brtek, M. D. (2000). The

nature and outcomes of work: A replication and

JENNIFER D. NAHRGANG is a doctoral student in management and organizational be-

havior at the Eli Broad Graduate School of Management at Michigan State University.

She graduated with a BS in business administration and an MBA from Oklahoma State

University. Her research interests focus on leadership, work design, and teams. Prior to

pursuing her PhD in management, she was a financial services supervisor for Exxon

Mobil Corporation.


23/24


extension of interdisciplinary work-design re-

search. Journal of Applied Psychology, 85,

860868.

Fogel, L. J. (1967). Human information processing. En-

glewood Cliffs, NJ: Prentice-Hall.

Fretz, B. R., & Leong, F. T. (1982). Vocational behavior

and career development, 1981: A review. Journal

of Vocational Behavior, 21, 123163.

Fried, Y., & Ferris, G. R. (1987). The validity of the Job

Characteristics Model: A review and meta-analysis.

Personnel Psychology, 40, 287322.

Gilbreth, F. B. (1911). Motion study: A method for in-

creasing efficiency of the workman. New York: Van

Nostrand.

Grandjean, E. (1980). Fitting the task to the man: An

ergonomic approach (3rd ed.). London: Taylor &

Francis.

Hackman, J. R., & Lawler, E. E. (1971). Employee reac-

tions to job characteristics. Journal of Applied Psy-chology, 55, 259286.

Hackman, J. R., & Oldham, G. R. (1975). Development

of the Job Diagnostic Survey. Journal of Applied

Psychology, 60, 159170.

Hackman, J. R., & Oldham, G. R. (1976). Motivation

through the design of work: Test of a theory. Orga-

nizational Behavior & Human Decision Processes,

16, 250279.

Hackman, J. R., & Oldham, G. R. (1980). Work re-

design. Reading, MA: Addison-Wesley.

Hackman, J. R., Pearce, J. L., & Wolfe, J. C. (1978). Ef-

fects of changes in job characteristics on work atti-

tudes and behaviors: A naturally occurring quasi-

experiment. Organizational Behavior & Human

Decision Processes, 21, 289304.

Hammer, M., & Champy, J. (1993). Reengineering the

corporation: A manifesto for business revolution.

New York: Harper Business.

Hammer, M., & Stanton, S. A. (1995). The reengineer-

ing revolution. New York: Harper Business.

Harvey, R. J. (1991). Job analysis. In M. D. Dunnette &

L. M. Hough (Eds.), Handbook of industrial and or-

ganizational psychology (2nd ed.) (pp. 71163).

Palo Alto, CA: Consulting Psychologists Press.

Herzberg, F. (1966). Work and the nature of man.

Cleveland, OH: World Publishing Company.

Herzberg, F. (1968, January-February). One more time:

How do you motivate employees? Harvard Busi-

ness Review, 48, 5362.

Hollenbeck, J. R., DeRue, D. S., & Guzzo, R. (2004).

Bridging the gap between I/O research and HR

practice: Improving team composition, team train-

ing and team task design. Human Resource Man-

agement, 43, 353366.

Homans, G. (1950). The human group. New York: Har-

court Brace Jovonovich.

Hulin, C. L., & Blood, M. R. (1968). Job enlargement,

individual differences, and worker responses. Psy-

chological Bulletin, 69, 4155.

Jamal, M. (1984). Job stress and job performance

controversy: An empirical assessment. Organiza-

tional Behavior & Human Decision Processes, 33,

121.

Johns, G., Xie, J. L., & Fang, Y. (1992). Mediating and

moderating effects in job design. Journal of Man-

agement, 18, 657676.

Juran, J. M., & Gryna, F. M. (Eds). (1988). Jurans qual-

ity control handbook. New York: McGraw-Hill.

Lawler, E. E. (1969). Job design and employee motiva-

tion. Personnel Psychology, 22, 426435.

Lawler, E. E., Hackman, J. R., & Kaufman, S. (1973).

Effects of job redesign: A field experiment. Journal

of Applied Social Psychology, 3, 4962.

Le Blanc, P. M., de Jonge, J., de Rijk, A. E., &

Schaufeli, W. B. (2001). Well-being of intensive care

nurses (WEBIC): A job analytic approach. Journal

of Advanced Nursing, 36, 460470.

Lepak, D. P., & Snell, S. A. (1999). The human resource

architecture: Toward a theory of human capital allo-

cation and development. Academy of Management

Review, 24, 3148.

Loher, B. T., Noe, R. A., Moeller, N. L., & Fitzgerald, M.

P. (1985). A meta-analysis of the relation of job

characteristics to job satisfaction. Journal of Ap-

plied Psychology, 70, 280289.

Losey, M. (1998, March 15). HR comes of age. HR

Magazine, pp. 4453.

Martin, R., & Wall, T. D. (1989). Attentional demand

and cost responsibility as stressors in shopfloor

jobs. Academy of Management Journal, 32, 6986.

McCormick, E. J. (1976). Human factors in engineering

and design. New York: McGraw-Hill.

McCormick, E. J. (1979). Job analysis. New York: AMA-

COM.

Meister, D. (1971). Human factors: Theory and prac-

tice. New York: Wiley.

Morgeson, F. P., & Campion, M. A. (1997). Social and

cognitive sources of potential inaccuracy in job

analysis. Journal of Applied Psychology, 82,

627655.

Morgeson, F. P., & Campion, M. A. (2002). Minimizing

tradeoffs when redesigning work: Evidence from a


24/24


longitudinal quasi-experiment. Personnel Psychol-

ogy, 55, 589612.

Morgeson, F. P., & Campion, M. A. (2003). Work de-

sign. In W. C. Borman, D. R. Ilgen, & R. J. Klimoski

(Eds.), Handbook of psychology: Industrial and or-

ganizational psychology (pp. 423452). Hoboken,

NJ: John Wiley & Sons.

Oldham, G. R. (1976). The motivational strategies used

by supervisors: Relationships to effectiveness indi-

cators. Organizational Behavior & Human Decision

Processes, 15, 6686.

Oldham, G. R., Hackman, J. R., & Pearce, J. L. (1976).

Conditions under which employees respond posi-

tively to enriched work. Journal of Applied Psy-

chology, 61, 395403.

Parker, S. K. (1998). Enhancing role breadth self-effi-

cacy: The roles of job enrichment and other organi-

zational interventions. Journal of Applied Psychol-

ogy, 83, 835852.

Parker, S. K. (2003). Longitudinal effects of lean pro-

duction on employee outcomes and the mediating

role of work characteristics. Journal of Applied

Psychology, 88, 620634.

Parker, S. K., Chmiel, N., & Wall, T. D. (1997). Work

characteristics and employee well-being within a

context of strategic downsizing. Journal of Occu-

pational Health Psychology, 2, 289303.

Podsakoff, P. M., & Organ, D. W. (1986). Self-reports in

organizational research: Problems and prospects.

Journal of Management, 12, 531544.

Roberts, K. H., & Glick, W. (1981). The job characteris-

tics approach to task design: A critical review.


Roethlisberger, F. J

work redesign: eight obstacles and opportunities

Documents