Managerial Models of Organizational Effectiveness

Managerial Models of Organizational Effectiveness

Studies of organizational characteristics predictive of managerial judgments of overall effectiveness of subordinate organizational units are reported.^ The findings are interpreted within the framework of a hierarchical model of criteria of effective- ness. Measurable organizational characteristics serve as operational, short-run sub- stitutes for the more subjective, long-run ultimate criterion of organizational effective- ness. The general business and research and development models of criteria of organizational effectiveness show differences in ways consistent with the analyses of Woodward (1965) and Thompson (1967).

Concepts of organizational effectiveness are the basis of theories of management and orga- nization behavior and provide the rationale for normative theories of organization behavior and management practice. There is relatively little consensus, however, about the relevant dimensions or components of these concepts.

The wide variety of concepts of organiza- tional effectiveness found in theoretical formu- lations of organization behavior is a function of some inherent characteristics of “the criterion problem.” and makes comparison of organiza- tional research findings difficult. Thorndike (1949:120-124). differentiated three sets of criteria for research on employee selection: ultimate, intermediate, and immediate. The ultimate criterion is the achievement of a final goal, which is likely to be stated in broad terms not susceptible to practical assessment by out- side observers. Application of an ultimate criterion must be an evaluation by those best qualified to ascertain the final goal of the organization and its achievement, such as managers or officials responsible for sub- ordinate organizational units. Also, in practice. various midrange criteria (intermediate and immediate) that are relevant to the ultimate criterion and practical to apply tend to be used in short-run assessment of effectiveness. The determination of relevance typically is a rational

“•The studies reported here received support from the University of Minnesota’s Industrial Relations Center, Graduate School, and Graduate School of Business Administration and fronti corporate con- tributions to the Industrial Relations Center.

process because measures of the ultimate cri- terion are lacking. This rational process generates theoretical or conceptual models of organizational behavior, which demonstrate the instrumental relationships among and between variables proposed as midrange criteria and some concept of organizational effectiveness. The criteria increase in number as additional conceptual models of organizational behavior are developed.

Studies in recent years have been directed toward narrowing the number of relevant criteria of organizational effectiveness and in- vestigating empirical relationships among these criteria. Findings suggest some criteria relevant to the ultimate criterion of organizational ef- fectiveness, some models of empirical relation- ships among criterion measures appropriate to specific situations, and explanations of differ- ences among the models.

DIMENSIONS OF ORGANIZATIONAL EFFECTIVENESS

A wide range of organizational characteristics (structure, composition, behavior, productivity, etc.) have been proposed and applied as criteria of organizational effectiveness. Although each of these criteria is viewed as conceptually distinct, it seems reasonable that there is con- siderable empirical relationship among them. One study (Mahoney, 1967) investigated the empirical relationships among 114 character- istics that are often considered criteria of orga- nizational effectiveness. These characteristics can be viewed as midrange criteria of organiza- tional effectiveness potentially relevant to the

357

358 ADMINISTRATIVE SCIENCE QUARTERLY

ultimate criterion of effectiveness. A varied sample of 84 managers in 13 companies com- pleted questionnaires which solicited informa- tion about subordinate organizational units. Information was obtained for 283 organization units. Descriptive assessments of the organiza-

tion units were obtained using the 114 criterion characteristics, and a judgment about overall effectiveness of the unit was obtained at the same time. Eactor analysis of the 114 variable assessments suggested 24 relatively independent criterion dimensions (Table 1), which ac-

TABLE 1. DIMENSIONS OF ORGANIZATIONAL EFFECTIVENESS WITH STANDARDIZED REGRESSION

COEFFICIENTS.

Model

Dimension

Flexibility. Willingly tries out new ideas and suggestions, ready to tackle unusual problems.

Development. Personnel participate in training and development activities; high level of personnel competence and skill.

Cohesion. Lack of complaints and grievances; conflict among cliques within the organization.

Democratic supervision. Suhordinate participation in work decisions. Reliability. Meets objectives without necessity of foilow-up and checking. Selectivity. Doesn’t accept marginal employees rejected by other organizations. Diversity. Wide range of job responsibilities and personnel abilities within the organization.

Delegation. High degree of delegation by supervisors.

Bargaining. Rarely bargins with other organizations for favors and cooperation.

Emphasis on results. Results, output, and performance emphasized, not procedures.

Staffing. Personnel flexibility among assignments; development for promotion from witbin the organization.

Coordination. Coordinates and schedules activities with otber organizations, utilizes staff assistance.

Decetttralization. Work and procedural decisions delegated to lowest levels. Understanding. Organization philosophy, policy, directives understood and accepted by all.

Conflict. Little conflict with otber organization units about authority or failure to meet responsibilities.

Personnel planning. Performance not disrupted by personnel absences, turn- over, lost time.

Supervisory support. Supervisors support tbeir subordinates.

Planning. Operations planned and scheduled to avoid lost time; little time spent on minor crises.

Cooperation. Operations scheduled and coordinated witb otber organizations; rarely fails to meet responsibilities. Productivity-support-uiilization. Efficient performance; mutual support and respect of supervisors and subordinates; utilization of personnel skills and abilities.

Communication. Free flow of work information and communications within the organization.

Turnover. Little turnover from inability to do tbe job. Initiation. Initiates improvements in work metbods and operations. Supervisory control. Supervisors in control of progress of work,

Mnllipte correlation, R

General business

.07

.08

.07

.03

.13

.02

—.02

.04

—.05

.01

.06

—.08

—.01

—.08

—.09

—.04 —.12

.25

.11

Research and

development

– . 1 9

.23

—.00

.01

.27

—.16

—.03 —.09

.01

.14

.01

—.08

.19

—.04

—.01

—.06

—.04

.31

.33

.43 .12

07 01 09

03 76

—.27 .17 .12

.08

.79

Mahoney and Weitzel: MODELS OF EFFECTIVENESS 359

counted for 65 percent of the measured variance among the organizations, indicating that the 114 organizational characteristics may be mani- festations of 24 dimensions of midrange criteria of organizational effectiveness. Many of the 114 variables, although conceptually indepen- dent, were so closely related in the managerial measurements of the organizational units that they comprised a single dimension in account- ing for the empirical variance in observations.

GENERAL MODEL OF ORGANIZATIONAL EFFECTIVENESS

A second analysis was made to investigate the relationships of the 24 midrange criteria of organizational effectiveness to managerial judg- ments about ultimate overall effectiveness of the organizational units studied. This second analysis involved fitting a multiple-regression model using stepwise regression analysis to the 24 criterion measures and judgments of ultimate effectiveness of the 283 organizations. This model accounted for 58 percent of the variance in judgments of ultimate effectiveness (R = .74). This model used only four of the 24 criterion dimensions analyzed, and one of the four accounted for approximately the same

proportion of ultimate criterion variance as the combination of the remaining three dimensions. The regression model can be summarized as follows, with the standardized coefficients in parentheses:

Organizational effectiveness = (.42) productivity-support-utilization + (.22) planning + (.16) reliability + (.12) initiative

The dominant dimension in this model is productivity-support-utili2ation, a complex of characteristics of these three concepts. Al- though the concepts are separable, measures of them were so closely interrelated that they were treated as a single empirical dimension.

The result of the regression analysis yielded a deceptively simple model, which did not reveal the full complexity of relationships among the dimensions and judgments of organizational effectiveness. A close examination of the cor- relation matrix and of the results of successive stages in the stepwise regression analysis sug- gested the model of relationships of criteria of organizational effectiveness that follows and that is illustrated in Figure la.

a. General btjsiness model

OveroM effectiveness

Productivity

Reliability Initiation

Support

Utilization

Oevetopment

b. Research and development model

Overall effectiveness

Flexibility

Supervisory control

Reliability Cooperation Development

Flexibility

Cohesion Supervisory control

Coordination

Cooperation

Cohesion Supervisory support

FIGURE 1. DIAGRAM OF RELATIONSHIPS OF MIDRANGE CRITERIA OF ORGANIZATIONAL EFFECTIVE-

NESS TO OVERALL EFFECTIVENESS IN (A) GENERAL BUSINESS MODEL AND ( B ) RESEARCH AND DEVELOPMENT MODEL.

360

Model

ADMINISTRATIVE SCIENCE QUARTERLY

Efficient, productive performance is the pri- mary criterion of organizational effectiveness. Such performance is closely related to, and usually accompanied by. a high degree of man- power utilization achieved through job assign- ments that challenge and utilize the skills available, as well as manpower development resulting from formal training and reliance upon internal development of manpower re- sources. Supportive relationships within the organization also correlate with efficiency and appear to be a function of the cohesion obtained within the work force and the supervisory sup- port provided the work force.

Planning, another major criterion, refers to the degree to which the organization is able to cope with emergencies and to concentrate upon the primary goal. Other dimensions re- lated to planning but less uniquely predictive of overall effectiveness are: flexibility in chang- ing policy, practice, and behavior; degree of cooperation with related organizations; and supervisory control of activities and operations within the organization.

Finally, the degree of initiation of ideas and practices, and the degree of reliability (meeting objectives without the necessity of follow-up or checking) shown in organizational behavior appear as independent criteria of effectiveness. The secondary dimensions in the model tend to be descriptive of organizational behavior rather than output or performance; thus, they might be viewed as criteria of organizational capability for future output performance.

RESEARCH AND DEVELOPMENT MODEL OF ORGANIZATIONAL

EFFECTIVENESS

The regression model of criteria of organiza- tional effectiveness was derived from analysis of a very heterogeneous sample of organiza- tions.- A more homogeneous sample might be expected to account for a higher proportion of

^Tfae companies studied ranged in size from 175 to over 10.000 employees and were engaged in heavy manufacturing, insurance, wholesale trade, electronics, and finance. Organizational units studied within these companies were engaged in production, marketing, engineering, research, administrative services, and staff functions and ranged in size from 4 to over 1,000 employees.

variance in judgments of overall effectiveness, but the meager number of organizations in the homogeneous subsamples of the large sample made it difficult to draw firm conclusions about the models of effectiveness identified in analyses of these subsamples. Consequently, an analysis was made of dimensions of organizational ef- fectiveness within a homogeneous sample of research and development organizations.

Data for the investigation were obtained from four companies operating within a single in- dustry. Only research and development units of these companies were studied to ensure a more homogeneous sample. Research and de- velopment was a vital function in the industry, which was characterized by competition among product substitutes. Product life in the industry depended upon the rapidity with which product substitutes were developed and marketed. A product might have a long life if all of the basic processes of production are protected by patents, or a short life if substitutes are easily derived by competitors. The relative impor- tance of research and development in these companies is illustrated in the maintenance of separate research and development institutes in some instances and by vice-presidential ap- pointments for research and development in all instances.

Assessments of 114 criterion variables and judgments of ultimate effectiveness were ob- tained for 103 research and development orga- nizational units in the four companies. The data were obtained from 32 managers, each of whom supervised three or more subordinate units (e.g., division, department, section). These managers held positions ranging from vice president for research and development through second-level supervision. Each manager had at least one level of supervision reporting to him. Because of the relatively flat organiza- tion of research and development units in the four companies, most of the managers held second- and third-level positions within the hierarchy.

The linear multiple-regression model de- veloped in the earlier studies and utilizing all 24 dimensions accounted for 63 percent of the variance in judgments of effectiveness (R = .79 which is little different from the R = .76 ob- tained for the 24-dimension model with the general industry sample). Standardized regres-

Mahoney and Weitzel: MODELS OF EFFECTIVENESS 361

sion coefficients developed for tbe 24-dimension model applied to each sample are presented in Table 1 for comparison.

As in the general industry sample, most of the criteria judged to contribute to ultimate effectiveness were accounted for by relatively few of the 24 dimensions. The following model, with standardized regression coefficients in parentheses, was derived for the research and development sample of organizations using step- wise regression:

Organizational effectiveness = (.431) reliability -i-(.271) cooperation + (.193) development

A multiple correlation R = . 7 1 was obtained for the three-dimensional model as compared with R — .79 for all 24 dimensions. Only one of the three dimensions, reliability, also figured in the four-dimensional model derived for the general industry sample. Although cooperation and development were not among the four sig- nificant dimensions in the general industry model, they did appear in the expanded model (Figure la) .

TABLI-: 2. RANKED IMPORTANCE OF CRITERIA

OF ORGANIZATIONAL EFFECTIVENESS.

Dimension

Research and development

1 Reliability 2 Cooperation 7} Development 4 Turnover 5 Stk’ctivity 6 Flexibility

General business

1 P erf 0 rmance-suppo rt-utiliza tio n 2 Planning 3 Reliability 4 Initiation 5 Baigaiiiing 6 Supervisory support

Gem-rai business

3 12 10 22 IS 14

Research and develop men t

14 10 1

13 20 17

Table 2 presents a comparison between the results obtained from the general industry sample and the research and development sample. The six dimensions contributing most

to the prediction of judgments of effectiveness within each sample are listed in rank order of relative importance; the corresponding rank order of importance in the model fitted to the other sample is also indicated. Only reliability appears among the six dimensions in both samples; otherwise, a dimension of relatively high importance in the judgment of effective- ness in one sample typically has relatively little importance in the judgment of effectiveness in the other sample. The dimension, productivity- support-utilization, for examjjle, is critical in the judgment of effectiveness in the general indtistry sample, but contributes little in judg- ments of effectiveness in research and develop- ment organization units.

The simple three-dimensional model of cri- teria of organizational effectiveness derived for research and development organizations is very different from the four-dimensional model de- rived for the general sample of business orga- nizations. Criteria relating to output and pro- ductivity appear to be subordinated to criteria relating to behavioral characteristics of the organization units. The difference is less striking, however, when a more inclusive model based upon the relationships among all 24 criterion dimensions (Figure lb) is compared with the more inclusive model for the general industry sample (Figure la).

Model

Reliability is the primary criterion of orga- nizational effectiveness in research and develop- ment units. Productivity and planning are so closely related to reliable performance, that they account for no independent explanation of variance in judgments of ultimate effective- ness. Supportive relationships within the orga- nization are correlated with productive and reliable performance and appear to be a func- tion of the cohesion achieved within the work force and the supervisory support provided them. Supervisory control of activities and operations within the organization are as- sociated with planning and coping with emer- gencies.

A second important criterion is cooperation. Achievement of cooperation is accomplished through coordination of scheduling, and flexi- bility in changing and adjusting assignments.

A third criterion, which is significant in the

362 ADMINISTRATIVE SCIENCE QUARTERLY

effective research and development organiza- tion, is the continuing development of the skills of its members.

This general model of criteria of effectiveness of research and development organizations rein- forces several elements of the general business model. The criterion of reliability is critical in judgments of effectiveness in both settings. Managers in all of the settings studied describe the highly effective organizations as being more reliable than the less effective organiza- tions. Attributing reliability to an organiza- tion probably reflects repeated short-term ac- complishments over a long period of time and tends to be synonomous with overall effective- ness. Reliability is not synonymous with achieved productivity, however, and can be assessed independently of productivity. Plan- ning contributes to reliability in research and development organizations, although it con- stitutes an independent dimension in the gen- eral business model. Supportive and cohesive relationships within the organization appear to contribute to productive performance in both the research and development and the general business models. Likert (1963:29-34) also found that cohesive attitudes within the work force appeared to be necessary, but were not predictive of high productivity unless coupled with respect and support for the superior.

DISCUSSION Differences between the research and de-

velopment and general business models of orga- nizational effectiveness can be understood in terms of a hierarchical complex of criterion measures; with the ultimate criterion at the apex of the hierarchy. Typically, the ultimate criterion of organizational effectiveness refers to long-run goal achievement. This achieve- ment is difficult to measure in the short run, because the goals sought are broad and general and thus difficult to define in terms of specific measures. Managers and others therefore de- velop various levels of midrange criteria that are easier to measure and that can be applied instead of the ultimate criterion and justified on the basis of some relationship to the ultimate criterion. They may be justified as short-run prerequisites to achievement of the long-run ultimate criterion, as subcomponents of the ultimate criterion, or as associated predictors

of the ultimate criterion. They can be viewed as forming a hierarchy of criteria, those at the highest levels being most closely related to the ultimate criterion and those at the lower levels being least closely related to it.

From this standpoint, the models of criteria of organizational effectiveness in Figure 1 show a distinct difference between the general busi- ness model and the research and development model. General business managers tend to use productivity and efficient performance as close substitutes for the ultimate criterion of effec- tiveness. Other close substitutes for the ulti- mate criterion include planning, initiation and reliable performance. These high-order criteria refer to measures of output, whereas lower- order criteria tend to refer to characteristics of the organization climate, supervisory style, and organizational capacity for performance. The research and development managers, on the other hand, use cooperative behavior, staff development, and reliable performance as high- order criteria; and efficiency, productivity, and output behavior as lower-order criteria. Both groups of managers look to the same general set of midrange criteria, but they arrange these criteria in different hierarchical levels of rela- tionship to the ultimate criterion of overall organizational effectiveness.

One explanation for this difference is related to the different concepts of the ultimate cri- terion. One might infer from the general busi- ness model that the concept of the ultimate criterion parallels closely the economic criteria of profit, productivity, and efficiency. Orga- nizational and behavioral characteristics are perceived to be related to the economic criteria and somewhat predictive of them. The predic- tive relationships are not absolute, however, and the organizational and behavioral criteria are not perceived to be determining components of the general ultimate criterion, and thus oc- cupy a lower level in the hierarchy of criteria. An ultimate criterion which shows a more pro- fessionally oriented point of view might be inferred from the research and development model. Profit, productivity, and performance efficiency are viewed as second-order criteria related to first-order professional criteria, but not critical indicators of the ultimate criterion. Rather, organizational and behavioral char- acteristics of reliable performance, cooperative

Mahmey and Weitzel: MODELS OF EFFECTIVENESS 363

relationships, and level of professional com- petence and development better predict the ultimate criterion. This explanation is sup- ported by common stereotypes of prodtiction- oriented managers of line-operating depart- ments and of professionally oriented managers of staff and research operations. Some re- search into the value and attitude orientations of managers lends support both to these stereo- types and to this explanation (England, 1967).

An alternative explanation is related to basic differences between the work environment of the research and development organizations and the typical industrial organization, such as, the length of the production cycle and the predictability of profitable output. There is no standardized production cycle in the research organizations. Some research projects may con- tinue for years before achieving the desired output and even be terminated after years of work without having achieved this output; other projects may achieve the desired output within a relatively short time. The production cycle in research is linked to factors that make it difficult even to predict profitable output. The development of marketable products is a function of many variables exogenous to the research function, and lack of knowledge about these variables makes it difficult to predict or control profitable performance in the research unit. Research breakthroughs may occur as a consequence of a related discovery in basic research performed in another organization, or may be nullified by a prior discovery in a competing organization. On the other hand, business functions, such as production, market- ing, finance, accounting, administrative services, and transportation, tend to have far more stan- dardized performance cycles, and performance results are more directly attributable to the organizational unit. Therefore, research and development managers might share a common concept of the ultimate criterion of organiza- tional effectiveness with general business man- agers and be equally concerned about long-run profitable performance of the entire organiza- tion, yet employ different models of organiza- tional effectiveness criteria for judging orga- nizational units. The research and development manager might well reason that short-term efficiency of performance is less predictive of the ultimate contribution of research and de-

velopment than organizational characteristics such as the manner of performance (coopera- tive and reliable) and the quality of the orga- nizational staff.

This latter interpretation relates to the find- ings of Joan Woodward (1965:23-25, 71 ii) as she relates Burns’ classification of organiza- tions according to organizational procedures with her classification according to production system. Burns identified two systems of orga- nizational procedures: ( i ) mechanistic systems characterized by relatively traditional, rigid, and hierarchical procedures, and (2) organic systems characterized by relatively loose, adap- tive, and flexible systems. Woodward (1965: 65-67) identified three major categories of production systems: {1) production of integral products in unit or small batch amounts, (2) production of large batches of identical prod- ucts and mass production, and (J) process production in continuous or intermittent flows such as might be found in refining petroleum. Woodward reported that organic management systems predominated in the unit or small- batch and in the process production system, but mechanistic systems predominated in the large- batch and mass-production systems of produc- tion. Although Woodward’s study was limited to manufacturing firms, one might hypothesize that the same general relationship between technology and organization appears through- out industry. Data on technology and orga- nizational system were not collected in the present studies; however, it seems reasonable that the research and development organizations would most closely approximate Woodward’s unit or small-batch production system, whereas the general business organizations would fit the concept of large-batch and mass produc- tion systems. The differences noted between the work environments of general business and research and development also appear to be associated with Woodward’s differentiation by technology. Technology and related character- istics of the performance cycle again appear to be closely associated with organizational characteristics.

The comparative findings provided by these two models also support various propositions by James D. Thompson (1967:86-98) about organizational assessment. He distinguishes be- tween organizations that believe they have

364 ADINIINISTRATIVE SCIENCE QUARTERLY

complete knowledge about cause-effect relation- ships and those that believe their knowledge is incomplete. Using this distinction, one might reasonably argue that knowledge of cause-effect relationships is perceived to be far less com- plete in research and development organizations than in the usual production department. Thompson suggests the norm of efficiency will be used where the knowledge of cause-effect relationships is believed to be complete, and norms of organizational fitness, organizational rationality, and organizational ability to meet the expectations of interdependent units will be used where the knowledge of cause-effect rela- tionships is perceived to be incomplete.

The similarities between the general business and the research and development models are as important as the differences. The same 24 dimensions of organizational effectiveness were used to describe the two, and variance along these dimensions “explained” 58 percent of the variance in judgments of overall effectiveness in the general business organizations and 62 percent in the research and development orga- nizations. These 24 dimensions appear to provide a reasonable explanation of organiza- tional effectiveness in varied organizational settings. The relative importance of the di- mensions varies from one setting to another. but the set of dimensions appears equally rele- vant in each setting.

The studies reported here also suggest that a relatively small subset of criterion dimensions can be identified in each setting, which is as predictive of ultimate effectiveness as the full set of dimensions. Relationships between managerial assessments of organizations along the 24 dimensions are such that a small subset, probably 3-5, of the dimensions predicts most of the variance in the judgment of overall ef- fectiveness. Such a finding might be expected to emerge from the multiple regression analy- sis employed. This is not to say that other dimensions are not important in achieving orga- nizational effectiveness, only that they provide little unique contribution to the explanation of variance. More significantly, the predictive subset of dimensions varies from one organiza- tional setting to another, and this variation is plausible and meaningful in terms of identifiable differences in the settings. Research within different organizational settings should provide

further checks on the findings reported here and their interpretation.

The findings pose a number of questions for continued study. The differences between the two models of criteria of organizational effec- tiveness might be a consequence of differing personalities and values of the respective man- agers or of the differing technological environ- ments. Studies might be conducted with other organizational groups classified at varying distances along the Thompson or Woodward continua in order to observe the related varia- tions in models of criteria of organizational ef- fectiveness. Or methods might be developed to identify the models of organizational effec- tiveness held and applied by individual man- agers, and studies conducted to investigate relationships among these models, managerial values, and organizational environments.

Finally, the findings of this research provide a basis for relating evaluative research in managerial practice to multiple criteria. Much of this research in the past has, of necessity, utilized some global criterion of effectiveness. The findings here indicate that the global cri- terion of overall effectiveness is a function of a set of more specific dimensions, which varies from one setting to another. Research findings of studies using a global criterion of effective- ness in varied settings often are confusing, prob- ably because of variation in the composition of the implicit set of criterion dimensions applied. The various dimensions of organiza- tional effectiveness identified in this research, if used as multiple criteria in place of a global criterion, may yield more conclusive findings, which the manager might utilize as appropriate to his concept of organizational effectiveness.

Thomas A. Mahoney is a projessor oj in- dustrial relations and assistant director oj the industrial relations center at the University oj Minnesota at Minneapolis. William Weitzel is an assistant projessor oj industrial relations and psychology at the University oj Minnesota at Minneapolis.

REFERENCES

England. George W. 1967 Personal Value Systems and Expected Be-

havior of Pharmaceutical Managers. Work-

Mahoney and Weitzel: MODELS OF EFEECTIVENESS 365

ing paper. Industrial Relations Center, University of Minnesota.

Likert. Rensis 1961 New Patterns of Management. New York:

McGraw-Hill. Mahoney, Thomas A.

1967 ‘Managerial perceptions of organizational effectiveness.” Management Science, 14, 2 (October); B-76—B-91.

Thompson, James D. 1967 Organizations in Action. New York:

McGraw-Hill. Thorndike. R. L.

1949 Personnel Selection. New York: Wiley. Woodward, Joan

1965 Industrial Organization: Theory and Prac- tice. London: Oxford University Press.


Comments are closed.