College and Research Libraries An Organizational Paradigm for Effective Academic Libraries David W. Lewis The organizational structures of academic libraries are considered in light of the work of anum- ber of organizational theorists, and related work by librarians is reviewed. An organizational paradigm for effective academic libraries is developed based on this review. The paradigm is composed of five parts: a modified professional bureaucratic structure, flexible resource alloca- tion, the use of management information systems, the export of production functions, and the development of an organizational philosophy. begin with the assumption that the current organizational structures of academic libraries · will not accommodate the roles that will soon be expected of them and the academic librarians who inhabit them. Predictions of the coming paperless soci- ety have been around for some time, but it _is now clearly more than talk. The technol- ogy of scholarship is changing, and as Pa- tricia Battin has said, ''The effectiveness of new systems of access to scholarly re- sources will depend upon the cooperative efforts of the university community to identify and develop the substance of new structures to knowledge, a process which will demand new organizational capaci- ties in the university. " 1 I would add that libraries must find new organizational ca- pacities as well. Despite a history that includes some no- table examples of innovation and change, the more important truth about academic libraries is . that they are encumbered by record systems and by financial and orga- nizational structures, which discourage innovation and make it difficult for them to manage uncertainty. For many good reasons academic libraries have been, and in large part still are, inertial institutions. The large investment in card files makes change in these structures all but impos- sible, as the recent experience with AACR2 has shown. In addition, the bud- gets of most academic libraries allow little or no flexibility. This, along with a capital budgeting approach in most colleges and universities that does not consider depre- · dation or return on investment, makes the justification of capital investment difficult. If academic libraries are to be successful in the future they must change. Academic libraries have been primarily production organizations. Their product has been an organized collection of mostly print materials. The specific production technologies have changed, especially in the past twenty years, but the basic orga- nizational philosophy and operational principles have remained largely un- changed for the better part of a century . The organizational structure adopted by academic libraries reflects this production orientation. Most are bureaucratic hierar- chies or, to use another term, machine bu- reaucracies. In general, these organiza- tional structures have been effective. They have allowed for consistency and control and for considerable job specialization. They are suited to an environment where . David W. Lewis is Lehman librarian at Columbia University, New York, New York 10027. 337 338 College & Research Libraries coordination depends, as it does with li- braries' complex record structures, on standardization of work. This is changing. Service programs have become more important, and the stan- dardization of records is becoming more a national, and less a local, issue. As a result of the development of national biblio- graphic utilities, academic libraries are in- creasingly capable of purchasing rather then producing bibliographic control. This ability to jettison much of the produc- tion function may provide academic li- braries the opportunity to, as Battin has said, ''draw into an existing strength the talents and expertise of individuals newly committed to the management and provi- sion of scholarly information services and create an institutional capacity to re- invent the university in the electronic age. "2 This article will explore the organiza- tional structures that are available to aca- demiG: libraries. Fundamental to this effort is the beliet shared by Henry Mitzberg and others, that an organization must find the structure that is the best fit to its func- tion and environment and that this struc- ture needs to be consistently constructed. As Mitzberg states: It should become evident how all of its elements of structure and situation form themselves into a tightly knit, highly cohesive package . No one element determines the others; rather, all are locked together to form an integrated system. 3 The pieces must be put together so that they complement each other; they should be put together with purpose. To speak in William Ouchi' s terms, the organization should have a well-developed philoso- phy. In looking at the organizational is- sues that confront academic libraries, we will begin by examining the work of a number of organizational theorists, partic- ularly Mitzburg, Ouchi, Jay Galbraith, and Selwyn Becker and Duncan Neuhauser. This will provide a context for our discussion. We will then examine ·aca- demic libraries in light of this theory and · the work of some librarians who have ad- dressed organizational issues. The final section of the paper will propose a struc- ture and organizational philosophy that I believe will effectively serve academic li- July 1986 braries as they confront their new roles. MITZBERG: STRUCTURAL ALTERNATIVES AND ORGANIZATIONAL FIT Mitzberg begins his analysis of organi- zational structure by identifying five pri- mary components of an organization. 4 They are (1) the strategic apex, or the top management; (2) the operating core, which contains the people who do the ba- sic work of the organization; (3) the mid- dle line or the managers between the oper- ating core and the strategic apex; ( 4) the technostructure, which provides systems design, formal planning, and control; and (5) a support staff that provides indirect services, including everything from the mail room to legal services. Using these components as a basis, Mitzberg con- structs five organizational configurations. Each has its own strengths and weak- nesses; each is best suited for different tasks and different environments. 1. Simple structure. This configuration consists of a few managers in the strategic apex and an operating core. Notably miss- ing are the other elements, a middle line, a technostructure, and a support staff. Or- ganizations of this type tend to be small and are controlled and coordinated by di- rect supervision from the strategic apex. They tend to use simple technologies and purchase support services. Simple struc- tures are lean and flexible. Centralized control allows rapid and flexible innova- tion and an ability to operate in a complex and even hostile environment. But this configuration is not without its limita- tions. It is highly dependent on strong and skilled leadership and is not suited for growth beyond a size that can be managed by direct supervision. This configuration also has difficulty with efficient mass pro- duction. 2. Machine bureaucracy. This is Mitz- berg' s term for what most of us think of simply as bureaucracy. This configuration specializes in mass production and is most effective in a stable environment. It em- phasizes standardization of work and job specialization. To assure standardization this structure elaborates its administrative structures. This leads to a large middle line to, as Mitzberg puts it, "oversee the specialized work of the operating core and to keep the lid on conflicts that inevitably result from rigid departmentalization, as well from alienation that often goes with routine, circumscribed jobs."5 The tech- nostructure is expanded to develop the re- quired control mechanisms. In an effort to stablize its environment machine bu- reaucracies will internalize support func- tions in a large support staff. The prob- lems of this configuration are well known-it creates dull and repetitive work that alienates employees. Organiza- tions in this configuration seek control over their environment rather than adap- tion to it and are thus unlikely to be inno- vative. They have trouble using complex technologies that require a sophistication that is difficult to standardize or where the operating core consists of professional workers. While these problems exist, it is also true that machine bureaucracies can produce many products cheaply and effi- ciently and can provide stability by insu- lating themselves to resist shocks from the external environment. It is not without cause that this configuration is the most common in industrialized societies. 3. Professional bureaucracy. Law firms and colleges exemplify this configuration. It relies for coordination on the standard- ization and high level of skills of its opera- · tors, and many decisions, both opera- tional and strategic, are made by these operators. Professional bureaucracies tend to be decentralized and democractic for the professionals in the operating core. Because of this decentralization there is a small middle line and large spans of con- trol. The technostructure is also small be- cause many of its tasks are performed by the professional operators. The support staff, on the other hand, tends to be large in order to give the professionals as much· aid as possible. The strategic apex often . does not so much supervise the operating core as provide a link to the broader envi- ronment. Professional bureaucracies can operate effectively in very complex envi- ronment because they can develop and apply high levels of skill. They can adapt, but they have trouble with fundamental or revolutionary change. To quote Mitzburg, An Organizational Paradigm 339 ''This is not a structure to innovate but one to perfect what is already known.' ' 6 4. Divisional form. This configuration comes into existence most often when the organization's product is diverse. Organi- zational units are often market based. The units themselves may take on any config- uration, but they relate to the parent orga- nization by a control system that empha- sizes standardization of outputs, in most cases short-term profit. It is argued that this form allows for the distribution of en- trepreneurial risk across a large organiza- tion, that it allows for the centralization and more efficient use of the technostruc- ture and support staff, and that it is more adaptive than a machine bureaucracy. It is, however, dependant upon divisional goals that can be operationalized, quanti- fied, and compared as well as on perfor- mance that can be measured in the short term. These requirements can produce an incentive structure for division managers not in the best long-term interest of the or- ganization. Short-term profit taken at the cost of long-term growth and suboptimi- zation are the likely results. 4. Adhocracy. This configuration suits organizations that need to innovate in complex ways in complex environments. It is a fluid structure based on interacting project teams. As such it is difficult to de- scribe in traditional terms. Coordination and control come about through informal communication and mutual adjustment among the experts who make up the proj- ect teams. Power in an adhocracy is not based on authority or hierarchical position but rather on who has the expertise to best make a given decision. It is possible for ev- eryone in the organization to contribute to both strategic and operational decisions. · The direction in which the organization moves is determined from the bottom, based on how proje_cts develop; it can not be imposed from the top. This configura- tion, as Mitzberg states, "contradicts much of what we accept on faith in organizations-consistency in output, control by administrators, unity of com- mand, strategy emanating from the top.' ' 7 Adhocracies can be considered inefficient because they require slack resources and cannot be tightly controlled. It is a config- 340 College & Research Libraries uration that can put men on the moon, but would likely not fare well competing in a stable production environment. OUCHI: ORGANIZATIONAL CONTROL MECHANISMS ANDTHEORYZ Unlike Mitzberg, who looks at struc- . tures, Ouchi sees mechanisms of organi- zational control as the key to understand- ing where and why organizations are successful. But like Mitzberg, Ouchi is a contingency theorist; that is, he believes that different types of organizational con- trol work best for different tasks in differ- ent environments. In looking at effective organizational control mechanisms Ouchi and Raymond Price see two requirements. First, the mechanism must assure that members of the organization are moti- vated to pursue acceptable goals. Second, the members of the organization must have the information necessary to execute actions that will meet the goals. They ar- gue that there are three basic organiza- tional control mechanisms: markets, bu- reaucratic hierarchies, and clans.8 Markets use price as both a motivator, in that individuals seek to maximize per- sonal wealth, and as a way of conveying information. But while market mecha- nisms are often used, they are generally supplemented by the use of a bureaucratic hierarchy. This is because price, especially in an organizational context, can be diffi- cult to determine. Establishing the contri- bution made by different departments to a product, as anyone who has even briefly studied transfer pricing issues can a test, is far from a simple matter. Market theory states that for a market to be effective it must have perfect, or at least approxi- mately perfect, information. In most orga- nizations the cost of information makes a pure market impossible, thus there is of- ten a fall back to bureaucratic control. Bureaucratic hierarchies rely on the specification and monitoring of rules. The rules provide the required information on how members of the organization should act. The monitoring provides the motiva- tion, at least in situations where employ- ees accept employment in exchange for a willingness to allow their superiors to tell them what to do and monitor their perfor- July 1986 mance closely. Rules, though, tend to be rigid and crude devices for conveying in- formation; they simply cannot anticipate all possibile contingencies. To overcome this, in bureaucratic hierarchies, excep- tions are passed up to a superior who has the authority to resolve the exception. Dis- cretion is thus concentrated at the top of the organization. This creates a situation in which the employee is dependent on his or her superior. Bureaucratic hierar- chies typically use job specialization to make rules simpler and monitoring easier. The final organizational control mecha- nism identified by Ouchi and Price is the clan. A clan is a culturally homogeneous group whose members share both a com- mon goal and a common understanding of the best ways to pursue that goal. In the pure form, the individual is socialized so that his or her goals coincide with those of the organization; this provides the moti- vation. In a clan, individual decision mak- ing becomes almost instinctual; thus, the clan can function even in ambiguous situ- ations without the overriding structure re- quired by a bureaucratic hierarchy or the information structures required by a mar- ket. But a clan is dependent on a stable and homogeneous membership, which in an industrialized society is difficult to achieve. Ouchi and Price sum up the advantages these three mechanisms by stating Hierarchy is thus only one among the three mechanisms of social control over collective tasks. However, while markets require highly sophisticated price information for their opera- tion and clans require extreme homogeneity and stability, bureaucratic hierarchies can oper- ate with only partially committed, largely unso- cialized employees, working under conditions of extreme ambiguity. 9 They conclude that ''if a bureaucracy is the only social control mechanism that can · withstand the conditions accompanying industrialization, then we must agree . . . that bureaucratic hierarchy is here to stay. ''10 In looking at ways to mitigate the negative effects of bureaucratic hierarchies-task specialization and dependency-it is necessary, Ouchi and Price argue, to incorporate either aspects of the market or aspects of clan mecha- nisms. Unless the essential control mecha- nism is altered, they say, fundamental change in the organizati0n will not be pos- sible. Ouchi argues for the adoption of ele- ments of clan control to create what he calls the Theory Z organization. 11 The The- ory Z organization dispenses with exces- sive bureaucratic structures by developing mechanisms of socialization. It takes ad- vantage of the trust and subtlety that result from this socialization process to in- crease productivity. To achieve trust and subtlety, Theory Z organizations put a premium on long-term employment. They avoid excessive job specializatigp by rotating individuals through a number of organizational functions and promoting individuals slowly. Work is generally or- ganized in groups that, while their output is measured, are given much autonomy in structuring their tasks. Also of importance is a well-developed, clearly stated, and broadly accepted organizational philoso- phy. The organizational philosophy is the mechanism by which the individual is in- tegrated into the organization. By focus- ing on the goals of the organization it pro- vides the basis for individual and work-group decision making. The em- ployee of a Theory Z organization is moti- vated by the long-term commitment the organization is willing to make to him or her, and perhaps of more importance, by the organization's willingness to trust his or her judgment in a wide variety of situa- tions. Ouchi uses the term subtlety to indi- cate the sophistication of an information system that is based on a long-term knowledge of the task, the goals of the or- ganization, and the strengths and weak- nesses of one's peers. This combination allows for refinements that often lead to a high level of productivity. It should be clear that the Theory Z approach is not · without its costs. These organizations must generally avoid volatile markets and have been criticized for allowing slack re- sources. JAY GALBRAITH: CONFRONTING INCREASED UNCERTAINTY Galbraith examines organizations from a perspective that focuses on their information-processing ability. This ap- An Organizational Paradigm 341 proach is useful in exploring the reactions of organizations as they encounter uncer- tainty. Galbraith begins with the basic bu- reaucratic organization that operates by a set of rules and passes exceptions up a hi- erarchy. He states: The ability of an organization to successfully utilize coordination by goal setting, hierarchy, and rules depends on the combination of the frequency of exceptions and the capacity of the hierarchy to handle them. As the task uncer- tainty increases, the number of exceptions in- creases until the hierarchy is overloaded. Therefore, the organization must again take or- ganization design action. 12 In Galbraith's view, how an organization confronts increased uncertainty is one of the most important strategic decisions it can make. He suggests that there are two basic strategies. The first is to reduce the amount of information the organization needs to process. The second is to increase the capacity of the organization to process information. The first three of Galbraith's alternative approaches fall into the first strategy; the second two alternatives use the second strategy. 1. Environmental management. In this ap- proach the organization, rather than changing itself, attempts to modify its en- vironment. There are a number of mecha- nisms that can be used. Vertical integra- tion can lessen uncertainty in the supply of resources or in markets. A public rela- tions response may be effective in influ- encing the environment. Finally, the orga- nization can contract or withdraw from the uncertain environment. 2. Creation of slack resources. This ap- proach is simply to reduce the level of per- formance. More resources can be applied by building redundant systems, increas- ing inventories, or adding staff. These ad- ditional resources are called slack resources because they are used only at times of high demand, at other times they are slack. This approach has obvious costs, but it will lessen the stress on the organization · and will reduce the number of exceptions passed up the hierarchy. 3. Creation of self-contained tasks. This ap- proach changes the ways tasks are divided into subtasks with the aim of creating more self-contained units. One way of achieving this is the creation of units 342 College & Research Libraries based on output, geography, or client group, rather than on input, skill, or occu- pational categories. This reduces the out- put diversity and the complexity faced by a unit. A second method is to reduce the division of labor. The example cited by Galbraith is the need for computer pro- gramming expertise in three departments. In a functional organization a programmer would be hired and his or her time shared between the departments. In the self- contained approach, where there was not sufficient demand for a full-time program- mer, the professionals in the department would do their own programming. As Galbraith says, ''specialization is reduced but there is not the problem of scheduling the programmer's time across the three possible uses for it. " 13 4. Investment in vertical information sys- tems. This is the first of the two approaches that aim to increase the organiz;:1tion' s ability to process information. Such sys- tems typically consist of mechanisms for monitoring and adjusting organizational plans and targets. These systems are usu- ally called management information systems. They are almost always computer based and are supported by the principles of managerial accounting. The goal of this approach is to collect information on the operation of the organization, to summa- rize and analyze it, and to channel it to the appropriate places in the hierarchy so that operations can be continually modified to increase productivity. 5. Creation of lateral relations. The final approach decreases pressure on the hier- archy by employing a lateral decision pro- cess that can cut across lines of authority. Rather than bring the information to the point where the authority to make a deci- sion exists, this approach decentralizes the decision process by bringing the deci- sion to where the information exists. There are a number of such coordinating or liaison mechanisms. They can range from interdepartmental liaisons to task forces, or a matrix design for some or all of the organization. The major cost of this approach is the increased cost of manage- rial resources that must be devoted to group processes. There is also a risk that decisions taken at lower levels will reflect July 1986 local rather than organizational goals. Each of the five alternatives to confront- ing uncertainty has its costs, and an effec- tive organization will choose the alterna- tive or combination of alternatives that meet its changing requirements at the least cost. It is Galbraith's contention that if an organization makes no conscious choice a choice will be forced upon it. As he states: The organization must adopt at least one of the five strategies when faced with greater uncertainty. [Italics in the original.] If it does not consciously choose one of the five, then slack, reduced per- formance standards will happen automatically. The task information requirements and the ca- pacity of the organization to process informa- tion are always matched. 14 BECKER AND NEUHAUSER: VISIBILITY OF CONSEQUENCES Becker and Neuhauser create a complex contingency theory of organizational ef- fectiveness and structure that they call the "entrepreneurial theory." 15 Not all of the details of their work are of concern, but one concept is most useful in looking at ac- ademic libraries. They describe an organi- zational variable, which they call the visi- bility of consequences and define as, "the degree to which the owner of an organiza- tion can and does evaluate the costs of ob- taining a given level of goal attainment from a procedure-resource interaction.'' 16 In other words, visibility of consequences is high when the head of an organization can and does measure the cost of a pro- gram and that program's contribution to the organization's goals. Low visibility of consequences comes about when the head of an organization either cannot-or will not-assess the value of a program in these terms. One way to think of visibility of consequences is to consider the length of time between an action taken by an in- dividual and a reasonable evaluation of the success of that action and the quality of the individual's performance. The longer this period, the lower the visibility of con- sequences. Low visibility of consequences, Becker and Neuhauser claim, will lead to an in- crease in usurpation, that is the inappro- priate use of organizational resources ei- ther by individuals or by departments. In hierarchical bureaucracy, usurpation usu- ally takes a form that Becker and Neuhauser call ritualization of procedures. As they describe it, "Ritualization as a form of usurpation occurs when an opera- tor insists on certain procedures not nec- essarily in the best interest of the organiza- tion simply because they benefit him personally." 17 For example, when an op- erator insists on maintaining old proce- dures even when new and better ones are available, the operator remains the expert, does not have to be retrained, and gains security. But the organization is less effec- tive. Becker and Neuhauser also suggest that a high visibility of consequences will assist in conflict resolution within organiza- tions. As they state, "when there is a basis for conflict resolution (high visibility of consequences) ... resolution will be rapid, and, other things being equal, lev- els of conflict low." 18 Organizations that have a high visibility of consequences, they claim, are not only quicker to suggest alternatives when faced with environ- mental change but will also make the nec- essary adjustments to react to change. When the effects of actions in an organiza- tion are clear, adaption is easier. For all of these reasons Becker and Neuhauser hy- pothesize that visibility of consequences is positively related to organizational effi- ciency and that procedures that enhance visibility of consequences will enhance the efficiency of organizations. After anum- ber of field studies Becker and Neuhauser claim that there is substantial evidence to support their hypothesis. LIBRARIES IN LIGHT OF THE THEORY Organizational Structures If we look at academic libraries using Mitzberg's organizational typology, we see for the most part machine bureaucra- cies. As Barbara Moran states in her recent review: Academic libraries today are organized in m~ny different patterns, depending upon size, kind of institution, growth rate, geographic disper- sal, and available space. Regardless of the orga- An Organizational Paradigm 343 nizational pattern chosen, almost all academic libraries are structured in a hierarchical man- ner.19 Helen Howard comes to a similar conclu- sion after reviewing the literature, "Bu- reaucratic models of organization domi- nate most libraries. Few major chan~es have been reported in the literature. " 2 In one of the most exhaustive analyses of the organizational structure of a single aca- demic library, the Booz, Allen, and Hamil- ton study of the Columbia University Li- braries, we again see a hierarchical bureaucracy. 21 While the details of the structure recommended by Booz, Allen, and Hamilton differ in some respects from previous library models, it has all the marks of a hierarchical or machine bu- reaucracy: a large middle line, a strong technostructure (in this case the planning office) and a support staff. Yet there is a tension. Booz, Allen, and Hamilton rec- ommend a standing committee structure to allow librarians from the operating core to participate in decision making. The structure of the resources group has all the markings of a professional bureaucracy. Another example of the alternative pulls of the machine and professional bureau- cratic configurations can be seen in Gor- man's description of two reorganizations at the Library of the University of Illinois at Urbana-Champaign. 22 He describes one reorganization that introduced six new as- sistant directors-an increase in the mid- dle line that might be seen as a move to- ward a machine bureaucracy. The second reorganization created decentralized groups of professional librarians to carry out selection, cataloging, and reference work, while centralizing clerical and automation-based processing. This looks very much like a move toward a profes- sional bureaucracy. As these two exam- ples show, large academic libraries con- . tain both production and professional functions. Martell presents models for ac- ademic library organization much like Mitzberg' s professional bureaucracy; it is based on what he calls "client-centered work groups. " 23 He solves the profes- sional/production dilemma by placing the production segments of the organization in a support staff role. 344 College & Research Libraries Joseph McDonald looked at libraries in light of Mitzberg' s theory. 24 He concluded that the professional bureaucratic configu- ration was the one that most fitted li- braries. However, he notes a number of weaknesses that might effect this configu- ration's effectiveness in the library con- text. They are (1) coordination by skill standardization is a loose coordinating mechanism that might not be effective in complex organizations; (2) professionals may use their discretion to pursue their own concerns and not the organization's goals, and (3) because innovation requires cooperation and the rearrangement of ac- cepted professional practice, it is difficult to achieve in professional bureaucracies. 25 Smaller institutions, such as college li- braries, unlike their larger cousins, are likely to be configured as simple struc- tures. This allows well-administered col- lege libraries to create programs with a fo- cus and coherence impossible in most university libraries. The example of the Earlham College Library under Evan Far- ber comes quickly to mind. When we see college libraries as simple structures the reasons for the challenges and opportuni- ties of college librarianship discussed by William Moffett become clear. Moffett de- . scribes the overriding need for good staff, especially good directors, and the possi- bilities for achievement that exist in the college environment. 26 As Mitzberg would tell us, a well-run, simple structure can outmaneuver a bureaucracy anytime. It is possible to view some academic li- braries, especially large systems with many branches, as divisional structures. This is generally not the case. What is more likely is that they are hierarchical bu- reaucracies segmented by subject or loca- tion rather than a true divisional form where the divisions have a high degree of autonomy. Finally, there are aspects of library oper- ations that may look like adhocracy. Proj- . ect teams are used for some large under- takings, such as planning and implementing online catalogs, and com- mittees exist in most academic libraries in frightening numbers. But these decentral- ized decision structures make recommen- dations as often as policy, and rarely do they fully manage a project. Joanne Euster July 1986 and Peter Haikalis have described the ap- plication of what they call a matrix struc- ture, a common adhocracy mechanism, to the public services department of a medium-sized university .library. 27 This however looks more like a decentralized of functional responsibilities than a true matrix structure. Innovation and Organizational Structure Klaus Musmann also considers libraries in light of Mitzberg' s work. His conclu- sions are much like those reached above. 28 Musmann goes on to suggest that because of the need to innovate, libraries may be forced to adopt adhocracy as a organiza- tional structure. I believe this conclusion indicates a confusion over the nature of in- novation in academic libraries. It is clear that vital organizations, including aca- demic libraries, will change as their envi- ronments, especially their technological environments, change. But what change is adaption and what is innovation? In their recent article, Judy Reynolds and Jo Bell Whitlatch concluded, after consider- ing a number of definitions of the term in- novation, that ''most relevant to the· stu~y of innovation in libraries is Mohr's defini- · tion of innovation-the successful intro- duction into an applied situation of means or ends that are new to that situation. '' 29 This is more than semantics. The creation of a complex product is very different from adapting that product to a new niche. If we accept Reynolds and Whitlatch's con- clusion, it is Mohr's definition of innova- tion that we should apply to academic li- braries. When we look at organizational theory, we need to be careful not to take as models organizations that are designed to innovate in the first sense. Effective aca- demic libraries must be able to adapt; they must be able to acquire and put to use new technology, but they do not need to create the new technology. In one of the few systematic studies of innovation in academics Howard found a positive relation between innovation and complexity and a negative relation be- tween innovation and centralization, for- malization, and stratification. 30 Howard defined innovation more as adaption than as a creation. Her results suggest that pro- fessional bureaucracy as well as adhocracy may be a structure that will allow this type · of innovation. Howard's study confirms that machine bureaucracy is not a suitable structure for academic libraries in the cur- rent dynamic technological environment. Organizational Control Mechanisms The problems presented by bureaucratic control mechanisms have been clear to li- brarians for some time. Connie Dunlap states in her review of academic organiza- tional patterns: In many libraries bureaucratic organization is being replaced by a collegial system. Bureau- cratic organizations tend to produce conformity and generally stifle creativity. Participative sys- tems, on the other hand, generally produce staffs which are not only more interested in the whole library and are more productive, but also staffs which are more flexible and more readily adaptable to change. 31 . The work of Maurice Marchant tends to support Dunlap's view. Marchant exam- ined a number of academic libraries and concluded that management style, through the intervening variable of staff job satisfaction, had a clear influence on the quality of l~brary service as measured by faculty evaluations. This effect is ex- plained by increased motivation and by better upward communication both of which allow for improved ROlicy forma- tion by top management. 32 There are many who see the importance of moving toward participatory management, which might be seen as a form of clan control, in academic libraries. But as Nicholas Burckel has put it, "Certainly one conclu- sion is that while the enthusiasm for par- ticipatory management continues, the ex- perience of some institutions and the research of librarians have tempered some of the more extravagant claims of its early proponents.' ' 33 The reasons for some of the problems with participatory manage- ment in the library context become clear if we look at it in light of Ouchi' s work. For a number of reasons clan control will be dif- ficult to achieve. 1. Clan control relies on a strong organi- zational philosophy. This philosophy needs to be established and articulated by the top management of an organization, but given the turnover of top academic li- An Organizational Paradigm 345 brary administrators, this is difficult to achieve. 34 2. Few libraries are large enough to pro- vide both long-term employment and a reasonable chance of promotion. Most of- ten individuals achieve advancement by moving from library to library. 35 3. The requirement of specialized skills, either subject, language, or technical ex- pertise, makes job rotation and similar programs that provide a broad organiza- tional perspective difficult. 4. Professional socialization in librari- anship is weak. Unlike law or medicine where this socialization takes place as part of professional education programs, it is less likely in short, fragmented library school programs. Given these difficulties in developing and conveying a coherent organizational philosophy and developing staff loyalty, it is difficult to risk decentralized decision making. The clan mechanism of organiza- tional control relies on the organization's philosophy as the major mechanism for coordinating decisions. Without it, decen- tralized decisions may often not be in the best interest of the organization. Ouchi and other recent management theorists have argued against the use of market mechanisms, but in the library context I do not find their arguments com- pelling. Libraries have never operated in a market environment, so the extremes of the market do not exist in libraries. As I have argued, market mechanisms may provide academic libraries with a number of benefits including better and more ap- propriate services and increased ability to judge performance. To make this possible will require developing better manage- ment information systems, which could well provide a clear picture of the nature of operations. 36 Formula systems for allocat- ing materials budgets are implicitly based on a market mechanism, especially when use is a part of the formula. The largest im- pediment to the use of market mecha- nisms is the lack of flexible approaches to resource allocation. Richard Talbot de- scribes the traditional approach to library finance and budgeting: Just as it is only a modest exaggeration to claim that library finance is dominated by the 5 per- cent rule, it is also only a modest exag~eration 346 College & Research Libraries to claim that library budgeting-the allocation of funds received by the library-is dominated by the 60-30-10 rule: 60 percent of the library budget is for staff, 30 percent for acquisitions, and 10 percent for other costs. This pattern has persisted for so long that it approximates a his- torical norm .37 Libraries have generally been unwilling to alter their patterns of budgeting. The sta- tus quo may be safer than any change be- cause they have few reasons sounder than past practice upon which to base deci- sions. A market mechanism would provide in- centive to an individual or department by providing benefit to that individual or de- partment based on some objective mea- sure of performance. This might take the form of increased salary or more resources to support the department's programs. The use of market mechanisms requires systematic measures of cost and value, so it is possible to make the necessary com- parisons. The first half of the problem is, if not straightforward, at least possible. Methods for establishing and appropriat- ing costs are used in most large academic libraries, but usually only to establish overhead costs for outside funding agen- cies. 38 The second half of the problem is more problematic. I have suggested a mechanism that uses partial user subsi- dies and a free market for library ser- viees.39 There has been some good work on performance measures, and the theo- retical basis for the necessary information systems exists, but it rarely seems to be ap- plied in practice. 40 . Mechanisms for Confronting Uncertainty When we look at the means Galbraith has identified for dealing with unc·er- tainty, two can be ruled out from the start. Academic libraries are not in a strong posi- tion to manage their environment except to the extent that they wish to withdraw from it and become museums for the book. This is not generally considered to be an acceptable alternative. The creation of slack resources is also unacceptable be- cause of the costs involved. The other alternatives that Galbraith of- fers are useful to consider in the academic library context. The creation of self- • contained tasks would argue for the devel- July 1986 · opment of generalists, but with highly de- veloped professional skills. If adopted . they would require a continued commit- ment to a strong program of staff develop- ment. The adoption of this approach to dealing with uncertainty would suggest the use of the professional bureaucratic model. The investment in vertical information system, or as they are more commonly called, management information systems, has been discussed above. It is certainly a possible approach, but one that few aca- demic libraries have pursued with much vigor. Such . systems should provide a good picture of what is happening in the library. As Robert Runyon said in one of the first articles on the use of management information systems in academic libraries, "One of the persistent problems in aca- demic library planning and decision mak- ing is to obtain an accurate picture of ex- actly what is going on within the · library.' ' 41 One of the results of lack of good vertical information systems has been pointed out by Marchant: The tendency has been to evaluate inputs (such as funding, personnel, and collection size and growth) · and processes with the assumption that they will predict the adequacy of the out- puts. This weakness has led to overemphasis on internal efficiency and a lack of concern for the place of the library within its environ- ment.42 The problem is not that libraries lack sta- tistics, but rather that the statistics are not analyzed well and that they do not mea- sure outputs. Academic libraries often at- tempt to compensate for this by surveying user groups. While the intention is laud- able, the standard method, the question- naire, should be used with more care than is often shown. 43 One of the persistent problems faced by libraries as they try to construct management information sys- tems is the collection of data for manual systems. Sampling methodologies should be of some help44, but few librarians seem to possess the required skills. As the Uni- versity of Lancaster studies demonstrate, even in a manual system it is possible to collect data so that significant understand- ing can be gained and applied with posi- tive effect to library operations. 45 The ap- plication o~ automated systems should provide some relief. It is encouraging that some of the libraries with the most ad- vanced automation programs have devel- oped management information systems. 46 The creation of lateral relations is a strat- egy that academic libraries often use, as their committee structures attest . Com- mittees are generally seen not only as a way to increase lateral relations, but also as a way to allow for participation in policy formulation by non-administrative staff. Coordinator positions, especially for bib- liographic instruction or online searching are not rare. In some cases, for example, those like the one reported by Buster and Haikalis, these positions are quite formal. Liaisons between department libraries and central services such as cataloging are not unknown, and liaison programs with academic departments and other campus groups are quite common. The increase of lateral relations is a strategy for confront- ing uncertainty because it allows com- munication and decision making without involving the bureaucratic hierarchy. Un- less appropriate independent decisions are allowed, the effectiveness of this strat- egy will be severely hampered. Visibility of Consequences It seems clear that in almost all cases, the visibility of consequences in academic li- braries is low. In part, the low visibility of consequences is the result of being seen as once removed from the primary functions of the university-teaching and research. The result of the inability of librarians to demonstrate their contributions has been expressed by R. H. Orr: "Indeed, it can be argued that much of the difficulty experi- enced in getting support for libraries is at- tributable to the use of measures of quality and value that have little validity for nonli- brarians. " 47 It should be clear that it is not so much that libraianship is seen as unim- portant, rather it is difficult to understand the effect of the choices we make in pro- viding service. Is reference desk service of greater value than bibliographic instruc- tion? Should we duplicate high use mate- . rials or purchase more less frequently used items? How much quality can we af- ford in our cataloging? And is any of this more important than keeping the building open untillO p.m. on Friday nights? That An Organizational Paradigm 347 these questions do not have generally shared answers is a sign of low visibility of consequences. Because it is difficult to say what the effects of actions will be, it is hard to determine how to react to changing sit- uations. Resolving conflict becomes diffi- cult, because no one can tell which side of an argument is correct. As most academic libraries are bureaucratic hierarchies, as predicted by Becker and Neuhauser, usur- pation by ritualization of procedures is common. Michael Buckland has applied the no- tion of a double feedback loop to library service; it is important to the context of visibility of consequences because it ex- plains how academic libraries can con- tinue to function with a low visibility of consequences. As Buckland describes the double feedback loop: These two feedback mechanisms are substan- tially..independent of each other in important ways. The librarian' s action does not depend on response by the user; and the user's action does not depend on response by the librarian. Further, since library services are normally free , the library's income does not depend directly on the level of demand . Reduced demand, therefore, does not weaken the library as it would a business, where a drop in demand would redu.ce sales and, therefore, income. Quite the reverse, a reduces demand for free · services reduced the pressure, leaving the exist- ing resources more adequate to cope with the remaining demand. 48 Buckland is saying that because the users in academic libraries are rarely the same people who decide levels of funding or al- locations, the dissatisfaction they express has little effect on the actions of librarians. In fact, because increased resources rarely follow increased use, it is not in the best interest of librarians to generate use, as it will only make the system more congested and raise levels of frustration. Until man- agement information systems are in place that can measure library performance and budget systems are flexible enough to re- spond to changes in demand, the double feedback loop will remain, as will the low quality of many library services. As Buck- land states, "This [the double feedback loop] helps to explain a noteworthy cyber- netic aspect of library service: Library ser- vices can survive with remarkable stability even in the absence of effective library 348 College & Research Libraries management. Survive, that is, not ex- cel. "49 AN ORGANIZATIONAL PARADIGM If academic libraries are to be adaptive, if they are to be innovative in the sense of creatively using the new technologies, they must first be innovative in their orga- nizational structures. As Rosabeth Kanter states, "Indeed, it is by now a virtual tru- ism that if technological innovation turns far ahead of complementary social and or- ganizational innovation, its use in practice can be either dysfunctional or negligi- ble. ''50 The information technology exists. We have bibliographic utilities, integrated online catalog systems, online database searching, microcomputers, and laser disks. The challenge before us is to build the organizational structures which will allow us to use this technology creatively. The structure and organizational philos- ophy that is suggested by the theory and research discussed above, at least to this author, may be divided into five compo- . nents. These components will need to be applied differently in different situations, but taken together they should provide a framework for effective academic li- braries. The components are: 1. The use of a modified professional bureaucratic configuration as a model for organizational structure. 2. The creation of flexible resource allo- cation mechanisms. 3. The use of management information systems to monitor performance. 4. The reduction of the production functions performed within the organiza- tion. 5. The creation and articulation of a well-developed, detailed organizational philosophy. It is important to remember that these components should all be used together. It is the package that is important, not one · piece of it. Each component will be exam- ined in turn. The Modified Professional Bu~eaucratic Configuration The first requirement for an effective ac- ademic library is an ability to recruit and maintain a skilled and motivated staff. As July 1986 Battin has said, ''the quality of the library staff during the next decade will be more important to the future health and vitality of the university than the quality of the in- . structional and administrative staff. " 51 To attract and keep highly skilled staff moti- vated will require an organization that al- lows and encourages professional behav- ior. Even if academic salaries were to become and remain competitive, this would still be the most critical factor. Mar- chant has put it well: ''If people are hired with skills being demanded, the skills must be used. ''52 The professional bureau- cratic configuration allows professional discretion and the autonomy without which skilled staff will become frustrated and will not be motivated to achieve high levels of performance. A compensation system that allows recognition of profes- sional as well as administrative achieve- ment, such as the one at Columbia Univer- sity, described by Frederick Duda, will encourage the development of profes- sional practice and skills. 53 As part of the adoption of the profes- sional bureaucratic configuration, the middle line should be reduced and as many strictly administrative positions as possible should be eliminated. This would mean large spans of control for the few re- maining administrators. But because many decisions will be made at the operat- ing core this should not be a problem. First-line administrators should remain, at least in part, professional operators. This flattening of the organization should im- prove communication in both directions. The top management will be closer to the operating core and the intelligence it gath- ers through its daily operations, and the operating core will be closer to the organi- zational goals conveyed from top manage- ment. Support services should be sup- plied to relieve librarians of nonprofessional tasks. The technostruc- ture should be reduced, and replaced with a continuing commitment to train profes- sional staff in the use of new technologies, and planning and control methodologies. There is a legitimate concern that auton- omous, specialized professionals bound together by only collegial governance will be more interested in personal goals than . in the goals of the organization. Me- Donald, as stated above, noted the poten- tial for these problems. Martell's solution is to develop client-center work groups. The group may have within it experts, but the idea is to create a pool of talent, of say six to ten librarians, that can deal with a broad range of tasks. Martell suggests the formation of subject or clientele-based groups, which contain all the professional functions, from cataloging to reference. While this may work in some circum- stances, the particulars of the group are less important than the attempt to com- bine a broad range of tasks within it. Groups of six to ten are small by most standards, but in academic libraries, groups of this size should go a long way to breaking down barriers and reducing seg- mentation. That the group has within it many self-contained tasks, as Galbraith suggests, should make the groups and the organization more flexible. By focusing on the group as the primary unit within the organization, a balance to- ward the professional autonomy is cre- ated. By adding group goals backed by peer pressure to a common sense of pro- fessional practice, many of the concerns of organizational coordination should be re- solved. The knowledge of one's peers that comes with group work is one of the com- ponents of a clan control system. It moti- vates and encourages the good use of an individual's talents. The basic manage- ment of the groups should be collegial and participative. Outputs need to be moni- tored, but groups should not be told how to run their operations. As Kanter puts it, "One key to managing knowledge work- ers is to let them alone to use their knowl- edge.''54 Flexible Resource Allocation The information technology will allow for the creation of new programs, but it will require capital investment. At the · very least, a microcomputer should be on every librarian's desk. To build local files on any size will require access to main- frame computers. The capital investment required for a circulation system or online public catalog is the beginning of the trend, not a one-time expenditure. Aca- demic libraries will become more capital intensive institutions, and the 60-30-10 al- An Organizational Paradigm 349 location rule will have to be altered. Mak- ing the library a part of the wired univer- sity will require investment, and academic libraries will have to learn how to make these investments. Part of this process should be an attempt to encourage risk taking by creating pools of "venture capi- tal" within academic libraries; an atmo- sphere that allows entrepreneurial proj- ects at the lowest levels should be created. A second factor that will require more flexible resources allocation will be the de- velopment of professional work groups. · These groups must have a say in spending funds. The successful service programs that were developed over the last decade and a half, library instruction and online . searching, were programs that required little or no capital. The primary resource · that went into these programs was the time and energy of the librarians who de- veloped them. But the librarians con- trolled them.lt is not unreasonable to sug- gest that this was one reason for their success. As more capital investment be- comes necessary to develop service pro- grams, the professionals in the operating core will have to have an increased say in how this investment should be made. This is necessary both because they are in the best position to know what is needed, and because if programs and systems are im- posed from above the operators will have less motivation to see them succeed. It is important that librarians in the operating core be given the power to act. Effective management information systems and good internal accounting should assure that the money is generally well spent. This will require a major rethinking of allo- cation practice, but without enpowering the professionals in the operating core by providing project funding, little else mat- ters. Management Information Systems Management information systems will play two important functions. First, mea- suring the output of groups will provide feedback that will show how they are per- forming in light of institutional goals. Sec- ond, the visibility of consequences will be raised across the organization. It will be possible to observe which programs had what effect, and to reward performance 350 College & Research Libraries both in terms of the salaries of individuals and the resources made available to ,groups. It may seem that the measurement re- quired of a management information sys- tem runs counter to the discretion implied by the professional bureaucratic configu- ration. Kathleen Heim has countered this argument: For those top managers with an inclination to share power and control of decisions, the MIS [Management Information System] many actu- . ally make this process more palatable. Since an MIS allows monitoring of decisions, a manager inclined to share power may do so and continue to monitor the outcome with a capacity to deter- mine if a subordinate has acted in an acceptable manner. 55 If the double feedback loop can be broken with management information systems, then quality service will be noticed and should be rewarded. Contrary to what might be expected, management informa- tion systems will allow more professional . autonomy, not less. By concentrating on outputs, management information sys- tems should force groups to look outward toward their clientele. This should en- courage a broader exposure to the envi- ronment. Management information systems should also provide insights into opera- tions that will lead to more effective ser- vices. Despite much good quantitative work on the functioning of academic li- braries, little of this work seems to be a part of the day-to-day operations of li- braries. This need not be the case and . management information systems will be an important contributing factor to bring- ing about this change. This, in turn, should bring stronger arguments for sup- · port. A management information system should demonstrate what the library does in terms that are understandable to fun- ders. The demands on institutions of higher education are too great to rely on the platitudes about the ''heart of the uni- versity'' as the basis for financial support; By raising the visibility of consequences we should be able to move beyond the platitudes and begin to speak in terms of . · the library's contribution to the college or university. July 1986 Reduction of Production Functions This trend is well established. Binderies are already largely things of the past. Bib- liographic utilities are providing academic libraries an opportunity to spread the risk of large investments and to purchase a product that had previously been largely homemade. As was stated at the outset, : this change may allow academic libraries to escape from an organizational structure that has been dominated by the need to as- sure standard work procedures. By ex- porting much of the production part of what academic libraries do, there will be less conflict between the parts of the orga- nization that must be structured for pro- duction and the parts of the organization . that are structured to provide information services to the members of the academic community. The export of production functions will make academic library orga- nizations less schizophrenic. Organizational Philosophy In this consideration of organizational structure, the particulars of an organiza- tional philosophy are less important than the fact that it exists, that it is well devel- oped, fully articulated, and taken seri- ously by the organization. When well done it is the written expression of the cul- ture of the organization. To quote Ouchi: Some organizations do have a self-conscious awareness of the underlying values and beliefs that they represent. With such an explicit awareness, an organization's philosophy can be its most useful tool in uniting the activities of employees through a common understanding of goals and values. A philosophy provides a standard of responses to problems, explains why certain behaviors will be rewarded, and ac- counts for company image . . . . Moreover, this development of organizational culture can in part replace bureaucratic methods of giving or- ders and closely supervising workers, thus leading to both increased ~roductivity and sup- portive relations at work. . In the academic library context the devel- opment of an organizational philosophy may be particularly important. The issue of the inappropriate use of professional · discretion was raised above, to some ex- tent the use of work groups and increased visibility of consequences will provide a balance. But the tool that will best channel the energies of professional librarians will be an organizational philosophy. The development of an organizational philosophy in academic libraries will not be easy. Organizational philosophies gen- erally begin at the top, and as we noted above, academic library directors may not stay in one place long enough to leave their mark. In addition, advancement in academic librarianship often requires movement between institutions. Without stability, the cultural traditions of the or- ganization are lost. Despite these prob- lems, academic libraries have some ad- vantages. Many are part of institutions with long traditions and rich cultures. And while librarians' sense of profes- sional practice may be weak, it exists and can be built upon. Perhaps the most effec- tive approach to developing an organiza- tional philosophy in academic libraries is a broad-based planning program. Method- ologies to support such efforts have been developed by the Office of Management Studies of the Association of Research Li- braries; their existence should be seen as a major resource. By using staff involve- ment as a means of gaining a wide organi- zational view and support for the results, this method should overcome the en- demic problems academic libraries face in developing organizational philosophies. This work will not be easy, and in a world . of daily crisis the effort required may seem ill spent, but it will be the key to construct- ing effective academic libraries. A FINAL WORD When put together all of this comes to a An Organizational Paradigm 351 very simple principle: to provide aca- demic librarians the tools to achieve suc- cess. Herbert S. White put th_e problem this way: Libraries, to a greater extent than other political units, embrace or allow themselves to be co- erced into accepting objectives for which there · are no resources, no plan, and no hope of suc- cess. We tell our staff members to "do the best they can,'' thereby clearly absolving them of re- sponsibility but also depriving them of any hope of success and ultimate celebration. In other words, the objectives are gibberish, and nothing really matters. 57 :The paradigm I have suggested is de- . signed to encourage coordinated profes- . sional discretion applied toward the solu- tion of problems. It attempts to provide the necessary resources, and to measure the results. The times offer academic li- braries many opportunities, but they will be hard to grasp. Until academic librarians are provided with the tools with which to apply their talents and until the outcomes of their efforts are measured, they will not know much of White's "ultimate celebra- tion." To quote White again, "Nobody likes to lose all of the time, and if library workers perceive that they never had a chance in the first place, they will quite · understandably quit caring and trying.'' 58 Given the tools and encouragement to use their professional judgement, I predict the opposite result. One successful project will lead to another; we will find the en- ergy and imagination necessary to use and adapt the technology, and with that tech- nology we will make significant contribu- . tions to our institutions and to the ad- vanceme~t of knowledge. REFERENCES AND NOTES 1. Patricia Battin, "The Library: Center of the Restructured University," College & Research Libraries 45:174 (May 1984). 2. Ibid., p.176 . 3. Henry Mitzberg, ''Organization Design: Fashion or Fit?'' Harvard Business Review 59:105 Qan. /Feb. 1981). 4. See Mitzberg, ''Organization Design: Fashion or Fit?'' or The Structuring of Organizations: A Synthe- sis of the Research (Englewood Cliffs, N .J. : Prenctice-Hall, 1979). 5. Mitzberg, "Organization Design: Fashion or Fit?" p.108. 6. Ibid., p .109. 7. Ibid., p .111. 352 College & Research Libraries July 1986 8. William G. Ouchi, and Raymond L. Price, "Hierarchies, Clams, and Theory Z: A New Perspective on Organizational Development," Organizational Dynamics 7:35 (Aug. 1978). 9. Ibid., p .37. 10. Ibid. 11. William G. Ouchi, Theory Z: How American Business Can Meet the Japanese Challenge (New York: Avon, 1982). 12. Jay R. Galbraith, Organization Design (Reading, Mass .: Addison-Wesley, 1977) p.49 . 13 . Ibid., p.51. 14. Ibid., p.55. 15. Selwyn W. Becker and Duncan Neuhauser, The Efficient Organization (New York: Elsevier North- Holland, 1975). 16. Ibid., p.29. 17. Ibid., p.31. 18 . Ibid., p.37. 19. Barbara B. Moran, Academic Libraries: The Changing Knowledge Centers of Colleges and Universities, ASHE-ERIC Higher Education Research Report, no. 8 (Washington, D.C.: Assn. for the Study of Higher Education, 1984), p.31-32. 20 . Helen Howard, "Organization Theory and Its Application to Research Libraries," Library Trends 32:488 (Spring 1984). 21. Booz, Allen and Hamilton, Inc., Organization and Staffing of the Libraries of Columbia University: A Case Study (Westport, Conn.: Redgrave Information Resources Corp., 1973). 22. Michael Gorman, "Reorganization at the University of Illinois-Urbana/Champaign: A Case Study," Journal of Academic Librarianship 9:223-25 (Sept. 1983). · 23. Charles R. Martell, Jr ., The Client-Centered Academic Library: An Organizational Model, Contributions d in Librarianship and Information Science, no. 42 (Westport, Conn.: Greenwood, 1983). 24 . Joseph McDonald, "Organizational Structure and the Effectiveness of Information Organiza- tions," Drexel Library Quarterly 17:46-60 (Spring 1981). 25. Ibid., p.53. 26. William A. Moffett, "Reflections of a College Librarian: Looking for Life and Redemption This Side of ARL," College & Research Libraries 45:338-49 (Sept. 1984). 27. Joanne R. Euster and Peter D. Haikalis, "A Matrix Model of the Organization for a University Library Public Services Division,'' in Academic Libraries: Myths and Realities: Proceedings of the Third National Conference of the Association of College and Research Libraries, Suzanne C. Dodson and Gray L. Menges, eds. (Chicago: Assn. of College and Research Libraries, 1984), p.357-64. 28. Klaus Musmann, "The Diffusion of Innovation in Libraries: A Review of the Literature on Organi- zation Theory and Diffusion Research," Libri 32:268 (Dec. 1982). 29. Judy Reynolds and Jo Bell Whitlatch, ''Academic Library Services : The Literature of Innovation,'' College & Research Libraries 46:403 (Sept. 1985). ~ 30. Helen A. Howard, "Organizational Structure and Innovation in Academic Libraries," College & Research Libraries 42:425-34 (Sept. 1981). 31. Connie R. Dunlap, "Organizational Patterns in Academic Libraries, 1876-1976," College & Re- search Libraries 37:402 (Sept . 1976) . 32. Maurice P. Marchant, Participative Management in Academic Libraries, Contributions in Librarian- ship and Information Science, no. 16, (Westport, Conn.: Greenwood, 1976); and "Participative Mangement, Job Satisfaction, & Service," Library ]ournal107:782-84 (Apr. 15, 1982). Some of Mar- chant's work has been recently supported; see Dale Susan Bengston and Dorothy Shields, "A Test of Marchant's Predictive Formulas Involving Job Satisfaction," Journal of Academic Librarian- ship 11:88-92 (May 1985). 33. r;.Jicholas C. Burckel, "Participatory Management in Academic Libraries: A Review," College & Research Libraries 45:31 (Jan. 1984). 34. In 1973 Arthur McAnally and Robert Downs found that half of the directors of ARL libraries had changed in the past three years and suggested that the average tenure was no more than five or six years; see Arthur McAnally and Robert B. Downs, "The Changing Role of Directors of University Libraries," College & Research Libraries 34:103-25 (Mar. 1973). Metz found a somewhat longer, 8.1- year mean "survival" and a 10.9-year tenure for the directors of the large libraries in his sample; see Paul Metz, ''Administrative Succession in the Academic Library,'' College & Research Libraries 39:358-64 (Sept . 1978). In a recent study of college library directors in the Midwest, Mech found the average number of years as director was 6.89; see Terrence Mech, "Small College Library Di- rectors of the Midwest," Journal of Academic Librarianship 11:8-13 (Mar. 1985). 35. Barbara Moran found that about half of the directors in her study were_ external candidates. The An Organizational Paradigm 353 proportion for men, who tend to achieve directorships in larger institutions, was more than 70 percent external candidates; see Barbara Moran, "Career Patterns of Academic Library Adminis- trators," College & Research Libraries 44:338 (Sept. 1983). Ronald Karr, in his study of directors of major libraries, both academic and public, found that the directors had worked at an average of four different libraries; see Ronald Dale Karr, "Becoming a Library Director," Library Journal 108:343-46 (Feb. 15, 1983) . 36. David W. Lewis, "Bringing the Market to Libraries." Journal of Academic Librarianship 10:73-76 (May 1984). 37. Richard J. Talbot, ''Financing the Academic Library,'' in Priorities for Academic Libraries, Thomas J. Galvin and Beverly P . Lynch, eds. New Directions for Higher Education, no. 39, (San Francisco: Jossey-Bass, 1982), p.38. 38. See, for example, Peter F. Rousmaniere, Elaine F. Ciarkowski, and Nathaniel Guid, "Bridging the Budget Gap: An Approach to Creating Fair User Charges," College & Research Libraries News 44:69-71 (Mar . 1983). 39. Lewis, " Bring the Market to Libraries," p.74. 40 . For work on performance measures see Douglas Zweizig and Eleanor Jo Rodger, Output Measures for Public Libraries: A Manual of Standardized Procedures, (Chicago: American Library Assn., 1982) and Paul B. Kantor, Objective Performance Measures for Academic and Research Libraries (Washington, D. C.: Assn. of Research Libraries, 1984). The most comprehensive work on management informa- tion systems in libraries is Michael R. W. Bommer and Ronald W . Chorba, Decision Making for Li- brary Management, (White Plains, N.Y.: Knowledge Industry, 1982). Also of importance on this subject is Clinic on Library Applications of Data Processing: Library Automation as a Source of Manage- ment Information, F. Wilfrid Lancaster, ed . (Urbana-Champaign, Ill .: Graduate School of Library and Information Science, 1983). 41 . Robert S. Runyon, "Towards the Development of a Library Management Information System," College & Research Libraries 42:539 (Nov. 1981). 42. Marchant, Participative Management in Academic Libraries, p .113. 43. For descriptions of the problems in using questionnaires in a library context see Michael Halperin, "Determining User Preferences for Information Services," Drexel Library Quarterly 17:88-98 (Spring 1981) and Abraham Bookstein, "Questionnaire Research in a Library Setting," Journal of Academic Librarianship 11:24-28 (Mar. 1985). 44. See, for example, Michael Halperin, "Cluster Sampling Reference Transactions," RQ 17:328-33 (Summer 1978). 45. The best review of this work can be found in Michael K. Buckland, Book Availability and the Library User (New York: Pergamon, 1975). 46. See, for example, Ken Dowlin and Lynn Magrath, "Beyond the Numbers-A Decision Support System,'' in Clinic on Library Applications of Data Processing: Library Automation as a Source of Manage- ment Information, F. Wilfrid Lancaster, ed. (Urbana-Champaign, Ill.: Graduate School of Library and Information Science, 1983), p.27-58. 47. R. H. Orr, "Measuring the Goodness of Library Service: A General Framework for Considering Quantitative Measures," Journal of Documentation 29:330 (Sept. 1973). 48. Michael K. Buckland, Library Services in Theory and Context, (New York: Pergamon, 1983), p.112. 49 . Ibid ., p .113. 50. Rosabeth Moss Kanter, The Change Masters : Innovation and Entrepreneurship in the American Corpora- tion, (New York: Simon & Schuster, 1983), p.20 . 51. Patricia Battin, "Developing University and Research Library Professionals: A Director's Perspec- tive," American Libraries 14:22 (Jan. 1983). 52. Maurice P. Marchant, ''The Library Professional: An Improved Model. If Library Schools Provide It, Will Libraries Use It or Abuse It?" fournpl of Academic Librarianship 10:194 (Sept. 1984). 53. Frederick Duda, "Columbia's Two-Track System," College & Research Libraries 41:295-304 (May 1980). 54. Kanter, The Change Masters, p.56 . 55. Kathleen M. Heim, "Organizational Considerations Relating to the Implementation and Use of Management Information Systems," in Clinic on Library Applications of Data Processing: Library Au- tomation as a Source of Management Information, F. Wilfrid Lancaster, ed. (Urbana-Champaign, Ill.: Graduate School of Library and Information Science, 1983), p.64 . 56. Ouchi, Theory Z, p.112. 57. Herbert S. White, "Participative Management Is the Answer, but What Was the Question?" Li- brary Journal110:63 (Aug . 1985). 58. Ibid. I