H. ,\v ' ' 111^

Organizational Effectiveness:An Alternative Perspective^

GREGORY H, GAERTNERS. RAMNARAYAN

Case Western Reserve University

A paradigm is proposed that views effectiveness as the ability of anorganization to account successfully for its outputs and operations lo itsvarious infernal and external constituencies. This framework can he usedto assess effectiveness by examining organizational outputs and processesin a critical and holistic way. A review of (he existing literatures onorganizational effectiveness and a case example are used to develop theproposed framework.

In the early 1970s, one of the major program em-phases of the Environmental Protection Agencywas to give grants to municipalities to constructsewage treatment facilities. This program con-sistently accounted for upwards of 70 percent of theagency's total funding until the 1982 fiscal yearbudget. In that budget, amid much controversy, thefunding for this program was slashed. Judgingfrom interviews and published documents, there islittle reason to believe thai major changes in theway the program operated had occurred to calldown either criticism or the budget knife. Did aprogram that apparently had operated successfullyfor more than a decade suddenly become ineffec-tive?

This example is far from unique in the federalsector or among organizations in general today.Moreover, existing literature on organizational ef-fectiveness gives few tools to explain how dramaticshifts in the evaluation of organizations ean occurwithout similarly drastic changes in the organiza-tion's activities.

This paper proposes a framework for under-

'The research for this paper was supported by the U.S. Officeof Personnel Management. Contract ffOPM 23-KO. The authorswould like to acknowledge the help received from Karen Gaert-ner, David Akinnusi, Richard Cogan, James Colhns, Toni Den-lon, Irene Devine, Allan Plash, and Juliann Spoth, who par-ticipated in a brainsiorming session to generate ideas for lookingat effectiveness. Also, thanks to Darlene Wolf for her patienieditorial assistance.

Standing transformations both in how organiza-tions do their work and how they account for thework they do. It is argued that effectiveness inorganizations is not a thing, or a goal, or acharacteristic of organizational outputs orbehaviors, but rather a state of relations within andamong relevant constituencies of the organization.An effective organization is one that is able tofashion accounts of itself and its activities in waysin which these constituencies find acceptable. Byimplication, effectiveness is not a state but rather aprocess; it is a characteristic of relations and notoutputs; it is negotiated rather than produced. Ef-fectiveness inheres only partly in bottom linefigures. A fuller apprehension of it requires anunderstanding of why and for whom this (or any)particular set of bottom line figures was produced,and of the institutional vision and logic that supportthis account of the organization to its constituents.The approach here is to view effectiveness as theability of an organization to account successfullyfor its outputs and operations to Us various internaland external constituencies.

Existing Literature

A review of the existing literature on organiza-tional effectiveness reveals that the different defini-tions and approaches can be characterized by twomajor dimensions: (1) Focus of the definition.

97

Some definitions focus on measures of terminaloutcomes, such as profitability, survival, or goal at-tainment. Others tend to be more concerned withorganizational processes and structure.s. (2) Intend-ed use of the concept. There are approaches thattend to be organization-specific. Others are intend-ed for a generality of organizations. The latterstudies aim for general propositions about eitheroutputs or organizational processes and structures.The former utilize the detail available to explainevents m a given organization or class of organiza-tions in a less generalizable way. These two dimen-sions, when cross-classified, result in four distincttypes of approaches, as proposed in Figure 1.

General Output Measures (Cell I)

One set of approaches included in Cell I of Figure1 uses the traditional accounting measures, such asproductivity, profit, or return on investment ascriteria for effectiveness (Price, 1968). The secondset of approaches focuses on organizational healthand survival as the ultimate organizational out-come. Some of the recent impetus for this approachhas come from authors whose theoretical perspec-tives are rooted in population ecology (Aldrich,1979; Hannan & Freeman, 1978; McKelvey, 1980).

These approaches, which rely solely on generalmeasures of output, labor under several difficulties.Different organizations produce different things, asdo even different subunits in the same organization.Metrics that can convert these outputs into com-parable units across or even within organizationsare not always available. Thus subunits, such aspersonnel and production control departments andpersonnel occupying staff and even managerial

positions, produce outputs that are not easily quan-tified and are only indirectly related to most ter-minal outputs.

Second, the multiplicity of outputs produced bythe organization results in their being differentlyvalued by different constituencies of the organiza-tion (Friedlander & Pickle, 1968; Scott, 1977).From different perspectives different views of theeffectiveness of a given organization are obtained.Under such circumstances, to decide, by legal ortheoretical fiat, which outputs really matter does adissservice to the complexity and diversity oforganizations. Because the general measures of out-puts have these problems in abstraction and ap-plicability, one alternative has been to focus not onoutputs of a general nature, but rather on the out-puts of particular organizations of interest.

Organization-Specific Output Measures (Cell II)

Goal-centered approaches to organizational ef-fectiveness can be considered in this cell. Withinthis orientation, effectiveness would be assessed byhow well the goals of the organization are beingachieved. Organizational goal attainment yieldsvaluable insights about the organization's characterand behavior (Etzioni, 1964; Perrow, 1970), be-cause serious goal setting represents an attempt atoptimization of potentially conflicting organiza-tional factors, in light of particular past and presentcircumstances and desired future. Thus, goal-cen-tered approaches provide a useful degree of detailand context often lacking in general output mea-sures, at the expense of some generality in findings.

But analysis of goal attainment for evaluatingorganizational effectiveness has several limitations.

Figure 1Approaches to Organizational Effectiveness

INTENDED USE Oh CONCEPTS

General Organ izat ion-Spcci lie

z

—. ^.—1

£• '^ o

C "^ ''^-^ a.

1: Represcnia[i\e Sciiools:1 radiiii.>nal Accounting

^1ea^ure'̂Population Ecology

Hi: Representative Schi'̂ ols:Leadership, Cieneral

ManaceiiientCoiitingencv ThcoiiesSystems Tlieory

II: Representative Schools:Goal Attainnieiii

IV: Repre^entativc Schools:Process-Orieiited

Qiialitative t:.vaiuation

98

First, goals for programs and organizations aredynamic and are likely to change over time, partlyas rejections of changing external circumsEancesand partly due to changes in the political makeup ofthe organization it.seif (Bardach, 1977; Kahn 1977).Second, there is the problem of interpreting the usesof goals m organizations. In some cases goals aretreated as window dressing, designed not to oricnithe behavior of organizational members, but ratherto provide oniy symbolic recognition to some con-stituency (Galbrailh. 1967; Hannan & Freeman,1977). In other cases they are seen not as actualtargets to be aimed for, but rather internal messageswithin the organization of what behavior is hopedtor (Granick, 1967). Third, Weick (1979) notes thatgoals in organizations frequently are inventions tosuit activity already performed—they arc or becometlie organization's means of restructuring a ration-ale for past activity.

One problem conimon to both orgam/aEion-spe-cifie and more general approaches to organizationaleffectiveness can easily be seen. WhaEever the ouE-puts finally arrived at, an understanding of effec-tiveness must inchide not only these outputs, bulalso an understanding of the factors that are associ-ated with their production, in the absence of suchunderstanding, any output measure of effectivenessis simple but barren—clear in its measurement, bulinadequate in utility. Unless theories of effec-tiveness reach back into the processes, structures,and intentions Ehat are associated with it, little canbe said about what \o do to improve the effective-ness of organizations. As a result, theoretical locushas shifted to organizational processes and struc-tures which either in the general case or in specificcases, can be shown to be associated with the effec-tive provision of outputs.

General Measures (Cell 111)

Much of the recent popular and academic liter-ature on management style (Blake & Mouton, 1964;Etzioni, 1964; McGregor, I960), leadership (Fiedler& Chemers, 1974), decision making (Kepner Sc 7"rc-goe, 1981), and organizational structure (Chandler,1976; Galbraith, 1973) focuses on generally effec-tive or contingently effective aspects of organiza-tions. These theories argue that for some or allorganizations, it is possible lo specify aspects o(organizational activity that are likely to improveorganizational effectiveness. For some of these.

there is the Hat statement of good ways to manageor organize. Increasingly, however, these flatstatements are discarded m fa\or of treatments thatlake into account relevant aspects of history, con-text, environment, and (he like (Kotter, 1978; Scoii,1977; Van de Ven & Ferry, 1980).

These views o\ organizaiional effectiveness haveproblems oJ their own. Firsi, to achie\c generalitythese theories rely on very abstract (heoreiical no-tions of dubious applicability (discussions of uncer-(auity, technology, and the "•fit"' between themprovide good examples) or Uiey relv heavily on adhoc bridging assumptions m applymg more generaltheories to any particular case. Second, ihesetheories frequently neglect the x'aried uses andmeanings that more generally defined constructshave in particular settings. Thus, management in-formaucin systems and computer technology mayhave Ihe theoretical referent of reducing uncerrain-Ey, but the praciical import (and meaning) ofredistributing power in existing organizanons(Gricar, 1979). Ciiven Ehe problems of defining andapplying generally effeclive .structures and puvcesses, some authors ha\e attempted to developmodels of process and sirucuire that are organiza-tion-specific, mainly operating as guides to diag-nosis and change in particular systems.

Organizalion-Speeilie Views iCell IV )

As represeniativc of the Cell IV approach, ihefocus is mainly on process-qualitative evaluations.The process-qualiiative evaluation has evohed as aresponse to the limitaiions of ihe outcome-orientedviews of effectiveness, pariicularly their difficuhies\n specifying generally agrecd-upon outputs mea-sured in common metric (Weiss & Rein, 1965). Thisapproach consists esscndally of information gather-ing to moiutor the unfolding of a program as it isgoing on, noting particularh those events ihat arecritical, collecting docunicnis. and obscr\ ing ihe ef-fects of ihc program on uistiiutions and individuals.This can provide the decisiv)n maker \Mth iiifurma-Eion about the varictv' ot" foinis the program is tak-ing in individual agencies oi subuniis, !nforniaiK>nthat nuiy be useful in undersianding (he impact ofihe program and in modifying policy.

Unfortiiiiateiy, the fciiuircs recommending thisapproach (flexibility, openness lo information,adapiability) arc the same features ihai encourage

criticisms of it as being too diffuse, not results-oriented, and having little prescriptive power.Moreover, theories of effective organizaiionaloperation, whether they are general or organiza-tion-specific, share a common set of problems.

The view that there arc "effective" structuresand processes rests on either extremely mixed andgenerally tenuous empirical support or on the firmbelief that some structures must be effective in somesense. Second and more importantly, these exami-nations of the efficacy of organizational structureor process iinplicitiy or explicitly assume that thereis a "real" measure of effectiveness, measured ornot. This belief is far from easy to confirm. Untilsome understanding of what the dependent variableis or means is reached, the tie of structure or proeessto it must remain unclear.

Summary of Current Approaches

Four major categories of approaches to organiza-tional effectiveness were defined by the cross-classification of the dimensions of intended use(general or organization-specific) and focus (out-comes or processes and structures). Each has some-thing to recommend it, and each has fostered goodand useful research. At the same time, each hasdistinctive disadvantages, partly inherent, and part-ly owing to hmitations in the state of relevanttheory and empirical results.

Approaches aiming for general statements aboutorganizational effectiveness in outcomes or pro-cesses gain scope and coverage at the expense ofdetailed applicability to given organizations. Thisgain in detail for orgamzation-specific approachesallows identification of the multiple and possiblyconflicting goals for the organization pressed bymultiple constituencies. These approaches also aremore able to unearth the multiple meanings anduses of organizational outputs and processes. Thesegains, however, tend not to be easily generalized.Approaches that focus solely on the outputs of theorganization tend not to reveal how to make organi-zations more effective. Approaches focusing onprocesses and structures have a greater level of ap-parent instrumental utiHty at the expense of assum-ing the effectiveness of these processes or struc-tures. A paradigm drawing on each of these theore-tical perspectives will be proposed.

An Example

The Municipal Contruction Program in the U.S.Environmental Protection Agency (EPA) is usedhere as an example for a review of the applicationof the four generic approaches to organizational ef-fectiveness. One of the measures of effectivenessused lor different programs in the federal sector isthe productivity index calculated by the Bureau ofLabor Statistics, which is calculated as the ratio ofthe sum of different weighted "outputs" of the pro-gram and the workyear "inputs"—a clear measureof general efficiency. The intended use of the indexacross offices and agencies is to get, in the ag-gregate, a measure of the productivity of the federalworkforce. The productivity index, arguably, canbe called a genera! measure of output (Cell I) for theprogram. Over a period of years, the index reportswhether the productivity has increased or decreasedand by what percentage. But the index has severallimitations. It ts far from inclusive of all the pro-gram activities; the outputs do not include factorssuch as quality and timeliness; the weights associ-ated with the outputs are of questionable accuracy;and the index makes little effort to correct forchanges in outputs or weights over time. Perhapsthe most important limitation of the index is in itsutility—from the figures, very little can be learnedas to why productivity has changed. The index ex-ists primarily for reporting, in an auditing sense, tothe outside world. It does not, in fact, have muchimportance in influencing the behavior of organiza-tional members.

Although the summary productivity index existsmainly for external consumption, the measures ofoutputs specific to the different categories in (heprogram are considered useful and relevant inter-nally (Cell II). In this case, it was observed that overa period of two years the number of categories andthe weights associated with the outputs changed.Thus these outputs tell about the reorientation ofgoals and provide a picture of the unit's activitiesmore closely in line with the unit's real situation.

For understanding the outputs and the changes inthem, an examination is needed, not of the outputsthemselves, but rather of the decision making andpriorities within the agency, which are behind theplanning and production of these outputs. Such ananalysis would utilize a perspective of agency-specific views of effective operation (Cell IV), It

100

was discovered in interviews with key decision mak-ers thai this particular unit had been criticizedwithin the agency for excessive reliance on gettingcontract dollars out and insufficient attention towhether the money was well spent. This criticismwas reflected in the increased attention to certainoutputs dealing with project completion.

None of the analyses discussed above would pro-duce an accurate, useful picture alone. They !ackthe consideration of the general framework tyingsuch decisions together, as represented by analysesfocusing in general considerations of effectiveoperation (Cell HI).

Program objectives for the municipal construc-tion program in the early years were found to retlectthe organizational priorities of EPA at that time.EPA's early history was one of urgent activity tomeet legislated program objectives^by and large,efficiency and cost effectiveness were criteria to beconsidered as secondary. As the early push for pro-gram objectives receded a bit, as oversight tromvarious external bodies increased. EPA's decisionmaking began to include consideration of the costsand cost effectiveness of its regulatory, enforce-ment, and construction activities. These changesfiltered down to its different subunits and pro-grams. Thus, the changes in the goals of the munici-pal construction program between two year.s can beviewed in the context of a changing notion of whateffective federal programs are about in EPA and ingovernment at large.

The case bears out that measures of effectivenessfor the four generic approaches have one facet incommon—they are accounts of those outputs andoperations, and not the outputs or operationsthemselves. There is fairly extensive literature sug-gesting that organizations tend to tailor the ac-counts of their performance in different ways, de-pending on whether the measurements and reportsare for external or internal consumption (Altheide& Johnson, 1980; Boland, 1982; Edelman, 1977;Hopwood, 1972; Meyer & Rowan, 1977; Wildav-sky, 1972), Thus it obviously is important torecognize that these accounts of outputs orbehaviors are different from the outputs or behav-iors themselves in .several respects.

They are not the activities in all their complexity,but rather interpretive summaries of them, screen-ing out irrelevant detail and highlighting importantparts of the activity. Thus, the letter accompanying

the data sent to the Bureau of Labor Statisticsstates:

We have. . .deleted some last year ou tpu t s ihat are nolonger priority i tems. In their place we have added<;evera! new ou tpu t s . We believe iheso delet ions andaddi t ions accuraiely reflect the agenc\"s producsivityand program thrusis for the current year.

Moreover, the reports are noi neutral construc-tions of events. They generally are constructed withan audience or purpose in mind. In rhe examplegiven, the purpose was partly one of self-justifica-tion, and the audience was at least the Fiureau oiLabor Statistics, and possibly other oversight agen-cies in government as well. Thus, a footnote to thedata states:

The new outputs (for the municipal construction pro-gram) reflect addnional and more mtensive analyticalprogram efforts.

It was felt within and without the agency that thiseffort has been insufficient in the past in this pro-gram and needed to be emphasized.

it is also important to note that the weights in theformula for the productivity index were modifiedfor the year that was reviewed. With the old for-mula, there would have been about a 30 percentdecrease in productivity, as against the 12 percentincrease reported with the new formula. It is notsuggested here that the new formula was utilized toshow better productivity—the program effort hadshifted in emphasis and the figures should retlectthis shift. It merely is noted that EPA was far fromindifferent to the effects of these figures andwanted to get the best possible hearing for the pro-gram. This was particularly so given the level ofcriticism that had been levied against the program.

Finally, these accounts of outputs and activitiesdiffer from the outputs and activities themselves inthat these accounts are created and received in thecontexts of the creators and receivers, respectively.Changes in the intentions atid circumstances of ei-ther sender or receiver change the character oi Iheaccount.

In summary, the various approaches to effec-tiveness would each seek different measures oi ef-fectiveness relying on different sources oi data.Moreover, each measure represents an accotmi bythe organization of itself and its activities. The dif-ferent accounts are selective interpretations ofevents, have different aims and audiences, and aresent and received in different contexts. Given thesedifferences in measures, data sources, and ac-

101

counts, how can effeciiveness he interpreted in anysingular and useful way?

The Paradigm

Perrow suggests:Insiead of slarting wiih a dcpcndcni variable that n apresumed public good and asking why we are gettingII or not getting if, we should ask "who is gettingwhat" lYom the organization or "effectivenes.s forwiiom?" The question presupposes a definition oforganization ihat is different from that assumed bymany schoob of thought, if we define organiza-tions. . .as intentional human constructions whereinpeople and groups within and without the organiza-tion compete for outputs of inlercst to them underconditions of unequal power, we have posed ihe issueof effeciiveness quite differently than in (he otherperspectives (1977, p. 101).

This political model of organizations lias beenably championed by numerous authors (Abell,1975; Crozier, 1964; Dalton. 1959; Pfeffer, 1978;Pleffer & Salancik, 1974; Zald, 1970). In thisperspective, the work ihat the organization doesand the way it conceives of its work form part ofboth the means and the ends in a complex bargain-ing process. In ihis process coalitions of actorslocated within and without the organizaiion forgeimplicit or explicit agreements on the strategy andtactics by which the organization's activities will begoverned. Thus, one way to define effectiveness inorganizations i.s not as a thing, or a goal, or acharacteristic of organizational outputs or behav-iors, but rather as a state of the relations within andamong these coalitions. .\n effective organization isone that is able to create accounts of itself and its

activities that relevant members of the dominanicoalition and the task environment find acceptablein light of these agreements.

The accounts may be for various purposes to var-ious audiences and for a variety of activities, in thepresent example, the municipal construction pro-gram was justifying its outputs to the Bureau ofLabor Statistics and other external groups in ageneral way, and also its attainment of goals withinthe agency. It was noted that these goals changed inresponse to specific criticisms of the program and tcchanges in general notions of effective governmen:operations. A way to characterize the accounts thaiorganizations and subunits make of themselves ispresented in Figure 2.

In the model the question of for whom the ac-count is constructed—its audience—is shown in thecolumns. Organizations account for themselves mquite different ways, internally and externally. Theformal structures, procedures, and even persotinelfrequently are different for these different au-diences. Thus, in the example of municipal con-struction, accounts of program activities were intended for two constituencies—external ones, in theform of the productivity index and the general com-mitment to increasing the cost effectiveness of pro-grams; and internal ones, in the program's specificoutput goals and changing those goals to aceountbetter for the costs and effects of the program.

The second major dimension of organizationalaccounting is based on the character of the activityaccounted for. Some organizational activities are inthe nature oi creating frameworks. Others are oper-ational or in the nature of performing within the

Figure 2Oreani/ation's Accounts of Itself

ALiDlhNCE- FOR ACCOUNI•xicnial inlenial

< o

•^ ^ •

: l ^

X

'C ^ 1 /

^ l i " " c::

[V: 1

•.^ c :

I .

^UDiriNt.. PROCL-SSlunicical construction:

t h e p i o d i i c i i v i t yi n d e x .

. t C l i l M A l l O N PROCESS

^UIntL'ln il ^nn^truciion-general i.oIl^lJcratillll^I'l cosl et k'ctivcness.

II: IMPLhMENTAI ION PROCESSMunicipal consiriiciu>n:

specilic outputs.

Ill: INTFCIRATION PROCESS

Municipal consiniciion:ihc decisions u\ changegoals to increaseanalytic capacity.

102

framework.The management decisions concerning planning,

staffing, acquisition and allocation of resources fallinto the category of framework creation. The sym-bolic governance of organizations and the politicalprocess of aggregating varied interests of organiza-tional parties into a sense of mission and com-municating that mission internally and externallyalso are framework setting. For the municipal con-struction program, these decisions included thegeneral and specific decisions to increase analyticcapacity in this program and programs like it in linewith organizational priorities and goals. By con-trast, performing within the framework relies onthese framework setting decisions and translatesthem into output oriented action. For the municipalconstruction program these outputs were measuredby the general productivity index and the speeifieoutputs that form the overall index.

This model ties rather directly to the earlierdiscussion of perspectives on organizational effec-tiveness. Accounting for outputs to external sourcesgenerally takes the form of general output measures(the productivity index). By contrast, internal ac-counting for outputs generally is of the form oiorganization-specific output measures (attainmentof specific output goals). The organization, hovv-ever, must account for itself not only in terms ofspecific outputs, but also in terms of its approach togeneral and specific decisions. Thus the decision toincrease the analytic capacity of the municipal con-struction program was an organization-specificdecision about effective operation that was inti-mately tied to effectiveness. This decision, however,was associated with a more genera! commitment tocost effective operations for external consumption.

Thus, each of the four perspectives on organiza-tional effectiveness can be seen to focus on mea-sures of a particular kind of activity intended for aparticular audience. One advantage of a more in-clusive perspective is that it allows discriminationamong these different measures and shows howthey are linked in the activities of a given organiza-tion.

The cross-ciassification of the dimensions of au-dience and activity leads to four categories oforganizational accounting. In the model an examplefor the Municipal Construction Program has beenpresented in each quadrant. The processes in quad-rants 1 to 4 have been referred to as auditing, im-

plementation, legitimation, and integration, respec-tively.

Accounting for external consumption (related tooutputs) is referred to here as "auditing." Thiscategory is by far the most frequently seen form oijustification activities-—-reports to Congress, regula-tory agencies, and so on. Such reports generallycontain mueh bottom line information in terms ofcountable, attributable outputs. Frequently theseindices are not those against which the organizationactually measures its activities internaiiy.

Performances that may appear quite satisfactoryrelative to public goals may be unsatisfactory fromthe perspective of what have become internally op-erative goals or the reverse. In fact, standardsagainst whieh performance is reported to the exter-nal environment may not have much meanmg with-in the organization. And because performancewithin the organization is not evaluated by these ex-ternal standards, they do not have much direct in-Huence on the behavior of the organizational ac-tors.

The implementation process involves self-analy-tic accounts and includes information on activity oroutput quality, value, or efficiency. These containdepartmental or individual performance evalua-tions. These accounts probably are neither morenor less likely to be truthful than accounts of out-puts produced for external audiences. They are.however, more likely io be influential day to dayand of a qualitatively different form than those in-tended for external consumption.

The third process covers the way in whieh theorganization accounts for itself in framework set-ting terms in its external constituents. This account-ing is referred to a legitimation. This fortn oi ac-count assists and/or seeks support for organiza-tional activities oi a framework setting sort, justify-ing or promising not a particular product but rathera particular role, asserting noi a thing done, but acapacity or need to be able to do that thing andother things like it.

The integration process refers to internal ac-counts of framework setting activities that are veryimportant within the organi/.ation, because throughthese activities the internal management of organi-zation is conducted. The work of integrating pro-grams, deciding which will be given resources andwhich will have resources taken away and the pro-cess that underlies these decisions—that is, the

103

operation of the organization's mternal polilicaleconomy—is accounted for in this way. Of the fouractivities, integration probably is the least frequent-ly accounted for m a formal way. It does not pro-duce many tangible outputs.

There are several different sorts of linkagesamong elements in the model. First, there are thelegitimation and auditing accounts that organiza-tions make to actors outside the organization. Thelinkage between these organizational accounts andenvironmental expectations often is associated withthe organization's justifications for past activity orattempts to secure future support. It also is associ-ated with attempts by actors in the environment tomonitor and direct organizational activities. Thislinkage not only serves to clarify the demands of theenvironment on the organization's outputs anddecision making, but it also serves as the main con-duit for the organization's attempts to manage itsenvironment (Aldrich & Pfeffer, 1976; Child, 1972;Pfeffer & Salancik, 1978). It is possible thatthrough missed communication, design or environ-mental or organizational change, the organization'saccounts and expectations of the environment willfail to correspond.

A second sort of linkage can be focused in eachquadrant of the model. This linkage relates workprocesses and activities to accounts of them in eachquadrant. Here again, when theory in use departsfrom the espoused theory (Argyris, 1970), there is ahigh likelihood that work activities will differ fromaccounts of them. These mismatches may resultfrom goal displacement (Merton, 1957) or fromdesign.

A third set of linkages relates the different kindsof accounts to one another. It is suspected that ac-counting related activities tend to follow a cycleeither from the upper left box (auditing) throughthe lower left box (legitimation) clock-wise, or inthe reverse sequence from legitimation to auditing,counterclockwise.

The first set of changes discussed in the municipalconstruction program between the two year.s beganin the legitimation phase. Criticism of EPA's andother regulatory agencies' neglect of cost effec-tiveness in program operations had led to a wide-spread push for regulatory reform. This was trans-lated as an internal consideration in the decision toincrease the analytic capacity and to change thegoals of the program (the integration phase), which

led to a reformulation of the program's internalgoals and outputs (the implementation phase).This, in turn, led to a new set of outputs to bereported to the Bureau of Labor Statistics (theauditing phase).

The more recent changes in the municipal con-struction program operated in the reverse fashion.Congress and the Administration, through the bud-getary process, made it clear that the constructionof municipal waste treatment facilities was not apriority any longer (the auditing phase). This wastranslated into new goals for the program internally(implementation) and new models for the effectiveoperation of the program (integration). It is far toosoon to tell what effect these changes have on themore general considerations of effective agencyoperations (legitimation).

it should be noted, however, that the accountsand processes in the paradigm need not be so closelyconnected. The mission of the organization e.\-pressed in legitimation processes may be very dif-ferent from the outputs for which the agency is heldaccountable. Such was argued to be the case for theHeadstart program (Deutscher, 1977). This discre-pancy may, but need not, be a stimulus for change.As Weick (1976) argued, this loose coupling pro-vides a source of organizational stability.

To summarize, in this paradigm four relativelydistinct—though interrelated^spheres of organiza-tional accounts have been proposed. These corre-spond to the four generic approaches to studies oforganizational effectiveness developed earlier in thepaper. Clearly, any comprehensive attempt to as-sess organizational effectiveness will consider bothoutput and process, in boLh internal and externalaccounts.

Assessing Organizational Effectiveness

The worth of the approach to effectiveness of-fered here probably is best seen in the new questionsit encourages. The paradigm proposed suggests twobroad classes of questions not generally asked inassessing the effectiveness of a particular organiza-tioti or subunit. First, what are the various accounisthat the organization makes of its activities? Whaiactivities are being accounted for and to whom?How are these accounts changing over time? Sec-ond, how do these accounts mesh with the expecta-tions of actors in the task environment, with the

104

processes the accounts are supposed to represent,and with other accounts? Outlined are some tenta-tive methodologies that may be helpful in pursuinganswers to these questions.

Accounts of terminal outputs prepared by the or-ganization for external audiences, auditing ac-counts, are a sensible place to begin in assessingorganizational effectiveness. For most organiza-tions these are available through printed reports oras outputs of ordinary accounting functions. Theymust, however, be carefully interpreted. First, onemust find out whether the outputs selected as criti-cal are of major interest to the organization, or tosome particular set of its constituents, or in fact tono one. Second, if informal or implicit goals andoutputs emerge in the course of data collection, at-tempts should be made to generate accounts ofthese as well. Third, one must be sensitive not onlyto changes in the levels of outputs, but to qualita-tive changes in the outputs themselves.

For accounts of terminal outputs generated forinternal use, implementation accounts, some ofthese questions are of lesser importance. Becausethe organization generally selects these outputs orgoals, they are less likely to be arbitrary or irrele-vant to the organization. These goals, however,may be of different importance in motivating mem-ber behavior, and they may change over time.

The proposed paradigm suggests two accounts oforganizational activities other" than those of ter-minal outputs for internal and external consump-tion just described. These include accounts of theintegration process, which covers decisions thatallocate resources and prioritize activities withinprograms and that reorient ongoing activities andprograms. Moreover, explanations for some ofthese specific decisions rest in the most generalassumptions and philosophies about what consti-tutes effective management; these represent ac-counts of the legitimation process.

Accounts of the auditing and implementationprocesses generally are available in the organiza-tion's printed reports or are available as outcomesof standard accounting practices. In fact, the majorbranches of accounting—financial and cost (ormanagement) accounting—are associated with ac-counting for external and internal purposes, respec-tively. Accounts of legitimation and integrationprocesses, by contrast, are not generally collected inmost organizations and probably would have to be

gathered through interviews, meetings, or otherface-to-face techniques. Accounts of the former,genera! views on effective operation, will be par-ticularly difficult to tease out. These views will in-volve questions of "ideas in good currency" andbroad managerial strategy. As an example, a broad-ly held view within the organization could be, "Inthe long run, no cost is too high for a clean environ-ment." Or it could be. "A clean environment is de-sirable; but it must be an economic proposition." Itcan be seen that organizational choices would be in-fluenced by which oi these ideas is in good curren-cy. These are unwritten criteria used by the organi-zation to test solutions. At choice points, elimina-tion of some alternatives takes place on the basis ofsuch assumptions and sentiments—certain possibili-ties and directions are not just considered becausethey do not fit with organizational sentiments ormission.

in sum, the assessment of organizational effec-tiveness rests not simply on how much of particularoutputs is being produced, but also on the decisionmaking that sets the framework in which the pro-duction of these outputs is carried out. These latteraccounts are not generally available. They usuallymust be gathered through face-to-face methods.

Effectiveness also rests on the extent to whichthese accounts mesh with the environment, with theprocesses they are supposed to represent, and witheach other. Assessing the linkages among thesevarious elements presents particularly difficultmethodological problems. Rather than speakingmore generally about the organization's ability topreserve or affect its environment or to integrate itsactivities internally, tracking specific issues anddecisions through the organization's decision mak-ing and production processes will generate the detailand interpretation necessary. Here Scriven's (1959)"modus operandi" approach and the techniques ofprocess oriented qualitative evaluation seem ap-propriate. Common to these approaches is an em-phasis on finding out how a particular decision orissues interpreted in various sectors and levels oi theorganization. This sort of information is most eas-ily collected through repeated interviews with offi-cials who had some part in handling these decisions.

To be sure, collection of data relating to accountsin each of the four quadrants suggested by theparadigm and the linkages between them is suffi-ciently difficult that for any large organization it

105

would be a massive undertaking. In most organiza-tions a sensible alternative would be to focus onparticular activities, programs, or subunits oi in-terest, it might seem that such intensive investiga-tion of only a few parts o!' an organization placessevere limitations on the generality of findings ob-tained. This objection does not seem particularlycompelling for several reasons. First, more globalcoverage that attends only to terminal outputsprobably provides less generally useful results.Although the specific outputs of an organization orsubunit might be unique, many of the managementand work processes it utilizes are common to manyorganizations. A less labor-intensive approach ofcollecting and tabulating unique organizational out-puts tells little about the more common tasks oforganizations and their members.

Second, perfect measures of effectiveness are notas significant as are changes in effectiveness andwhat caused or could cause them. Such knowledgeis furthered best by detailed knowledge of theorganizational processes that translate environmen-tal demands into various administrative and pro-duction processes and accounts of them. The para-

digm proposed is intended to provide just suchknowledge.

Discussion

The approach to assessing organizational effec-tiveness presented here has several advantages overexisting approaches. The model allows an examina-tion of both outputs and processes in a critical andholistic way, and it recognizes the roles of each andthe linkage between them as critical to effectiveness,Effectiveness in the model is seen less as an endstate than as a continuous process relating theorganization to its constituents. Finally, with thefocus on processes common to many organizationsrather than outputs unique to each of them, theresults obtained can be both plausible and useful.

To be sure, there are limitations to the approachsuggested. It is expensive in terms of time and ef-fort, and it does not produce global conclusionseasily. It can, however, produce rich and usefulassessments of organizational outputs and pro-cesses that do justice to the variety and complexityof organizational life.

References

Abell, P. Organizalionr, as bargaining and in_fluence systems.New York: Halsicd Press, 1975,

Aldrich, H. E. Organizanons and environment^. EnglewoodCliffs, N. J.: Prentice-Hall, Inc., 1979.

Aldrich, H. E., & Pfeffer, J. Environments of organi^ations-Annual Review of Sociology, 1976. 79-iO5.

Altheide, D- L., & Johnson, J. M. Bureaucratic propaganda.Boston: Allyn and Bacon, 1980,

Argyris, C. Intervention theory and method: A behavioralscience view. Reading, Mass.: Addison-Wesley PublishingCo.. 1970,

Bardach. E. The implementation game: What happens after ahill become.^ a law. Cambridge, Mass,: The MIT Press, 1977,

Blake, R. R.. & Mouton, J. S, The managerial grid. Houston,Tex.: Gulf Publishing Co., 1964.

Boland, R. J, Myth and technology in the American Accountingprofession- Journal of Management Studies, 1982, iO9-I27.

Chandler. A. D, The visible hand. Cambridge, Mass,: HarvardUniversity Press, 1976.

Child, J. Organization structure, environmenl, and performance—The role of strategic choice. Sociolo^iy. 1972, 6, 1-22,

Crozier, M, The bureaucratic phenomenon. Chicago: Universityof Chicago Press, 1964,

Dallon, M, Men who manaj^e- New York: John Wiley & Sons,Inc,, 1959,

Deutscher, 1, Toward avoiding the goal trap in evaluaiion re-search. In F. G, Caro (Ed,), Readings in evaluation research.New York: Russei Sage Foundation, 1977, 221-238.

Edelman. M. Political language: Words that succeed and policiesthat fail. New York: Academic Press, 1977,

Etzioni, A. Modern organizations. Englewood Cliffs, N. J-:Prentice-Hall, Inc., 1964,

Fiedler, F, E,, & Chemers, M, M, Leadership and effectivemanagement. Glenview, 111.: Scott, Foresman and Company,1974,

Friedlander, F., & Pickle, H. Components of effectiveness insmall organizations. Administrative Science Quarterly, 1968,13, 289-304,

Galbraith, J- K- Designing complex organizations. Reading,Mass.: .Addison-Wesley, 1973.

Galbraith, J, K, The new industrial stale. Boston: HoughionMifOin, 1967,

Granick, D. Soviet metal fabricating and economic develop-ment: Practice versus policy. Madison: The University ofWisconsin Press, 1967.

106

Gricar, B. G. Environmental imperatives created by governmentregulation: Predicting organizational response- Unpublisheddoctoral dissenation. Case Western Reserve University, 1979.

Hannan. M. T., & Freeman, J. Obstacles lo comparative studies.In P. S. Goodman & J. M. Pennings (Eds.), New perspectiveson organizational effectiveness. San Francisco: Jossey-Bass,1977. iO6-l3l.

Hannan, M. T., & Freeman, J. H. The population ecology of or-ganizations. In M. W, Meyer (Ed.), Environments and organi-zations. San Francisco; Jossey-Bass, 1978, 131-171.

Hopwood, A. G, An empirical study of the role of accountingdata in performance evaluation. Empirical research in ac-counting (supplement to ihc Journal of Accounting Research).1972, 10, 156-182.

Kahn, R. L. Organizational effectiveness: An overview. In P. S.Goodman & J. M. Pennings (Eds.), New perspectives on orga-nizational effectiveness. San Francisco: Jossey-Bass, 1977,235-248.

Kepner. C , & Tregoe, B. The new rational manager. Princeton,N. J.: Princeton Research press, 1981.

Kotter, J. P. Organizational dynamics: Diagnosis and interven-tion. Reading, Mass.: Addison-Wesley Publishing Co., 1978.

McGregor, D. The human side of enterprise. New York:McGraw-Hill, 1960.

McKelvey, B. Organizational speciation. In C. Pinder & LMoore (Eds.), Middle range theory and the study of organiza-tions. Boston: Martinus Nijhoff, 1980, 169-186.

Merlon, R. K. Social theory and soctal structure. Glencoe, 111.:Free Press, 1957.

Meyer, J. W. & Rowan, B. Institutionalized organizations: For-mal structure as myth and ceremony. American Journal ofSociology, 1977, 83, 340-363.

Perrow\ C, Organizational analysis: A sociological view. Bel-mom, Cai.: Wadsworth Publishing Co., Inc.. 1970.

Perrow. C. Three types of effectiveness studies. In P. S. Good-man & J. M- Pennings (Eds.), New per.<,pective.<, on organiza-tional effectiveness. San Francisco: Jossey-Bass. 1977, 96-105-

Pfeffer, J. Organizational design. Arlington Heighis, HI.: AHMPublishing Corporation, 1978.

Pieffer, J., & Saiancik, G. R. Organizational decision makmg asa political process: The case of a university budget.Administrative Science Quarterly. 1974, 19, 135-151.

Pfeffer, J., & Salancik, G, R, The external control of organiza-tions: A resource dependence perspective. New York: Harper& Row, 1978.

Price, J- L. Organizational effectivenes.s. Homewood, 111.wm, 1968.

Ir-

Scott, W. R. Effec[i^eness oi organizaiional effectivenessstudies. In P. S. Goodman & J. M. Penning^ (Hd'-.), Sew Per-spectives on organizational effectiveness. San Francisco:Jossey-Bass, 1977. 63-95.

Scriven, M. Explanation and prediction m evolutionary theory.Science. 1959, 130, 477-482.

Van de Ven, A, H., &. Ferry, I). L. Measuring and assessing or-ganizations. New York: Wiley, 1980.

Weick, K. E. Educational organizations as loosely coupled sys-tems. Administrative Science Quarterly. 1976, 2!, 1-19,

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

VVeiss, R. S.. & Rein, M. The evaluation of broad aim programs:A cau(ionary case and a moral. The Annals of AmericanAcademy of Political and Social Science, September 1969,133-142.

WMldavsky, A- B. The self-c\'aluaiing organization. Public .Ad-ministration Review. 32, 509-520.

Zald, M. N- Political economy: A framework for comparativeanalysis, ln M. N. Zald (Ed.), Power in organizations.Nashville. Tenn.: Vanderbilt University Press, 1970. 221-261.

Gregory H. Gaertner is Assistant Professor of Organiza-tional Behavior in the Weatherhead School of Manage-ment, Case Western Reserve University-

S- Ramnarayan is an Advanced Doetoral Student in theDepartment of Organizational Behavior, Case H'esternReserve Universitv.

107