College and Research Libraries JEFFREY A. RAFFEL From Economic to Political Analysis of Library Decision Making* In general, the more critical the decision, the less useful a cost-benefit analysis is to library decision makers. Political analysis is required, and Easton's conceptual framework is presented to suggest the utility of political analysis. A list of normative issues is derived from raising descriptive questions about the politics of university libraries. IN 1969 THE M.I.T. PREss published a new volume, Systematic Analysis of University Libraries: An Application of Cost-Benefi.t Analysis to the M.I.T. Li- braries, which might have signaled the entrance of economic analysis into the area of library decision making. As co- author of the book, I anxiously awaited the reviews I hoped would follow. 1 To date all reviews missed what I regard as the major point of the book: Although helpful, an economic analysis of a uni- versity (or public) library is insufficient because libraries operate as political sys- tems and thus improving libraries re- quires political analysis. The purpose of this paper is not only to argue that political analysis of uni- versity and public libraries should be undertaken in conjunction with eco- nomic analyses but also to apply a spe- cific theoretical framework and concept to university and public libraries. The improvement of libraries requires an expansion of analysis beyond technical discussions of procedural changes and per item costs to the broader utilization of social science theory and research. In the past two decades both econo- Jeffrey A. Raffel is with the Division of Urban Affairs, University of Delaware, Newark. 412 I mists and political scientists have ex- panded their field of inquiry. Econo- mists have become crucial figures in the analysis of governmental policy, espe- cially in the measurement and an·alysis of governmental effectiveness through the methodology of cost-benefit analy- sis.2 Political scientists have at the same time shifted their focus toward non- governmental institutions, with some analyzing what were once thought to be nonpolitical governmental institutions (e.g., schools) and others, nongovern- ments (e.g., private governments). As Mancur Olson has recently observed, the social science disciplines differ by their approaches and theoretical frameworks rather than by their subject matter. 3 Thus libraries, be they primarily public- ly or privately operated, are fair game for the frameworks of political scien- tists and economists. Each discipline includes an array of theoretical frameworks. The overall ap- 0 My frequent co-author, Robert Shishko, has tried to impart the essence of cost-benefit analy- ses to his audience by telling them about an economist who, when asked if he liked sex, re- plied immediately, "What are the alternatives?" I thank Bob Shishko for helping me to learn enough about economic analysis to criticize it, and I thank David Schulz and Daniel Rich for their insightful comments on earlier drafts of this paper. [_, • > proach of systems analysis stands out within each discipline as a fruitful way to improve libraries, specifically cost- benefit analysis in economics and Eas- tonian systems analysis in political sci- ence. EcoNOMIC ANALYSIS Brief Description "Basically, economic analysis is the study of choice: the allocation of scarce resources among alternative uses, and the distribution of outputs among al- ternative uses-that is, the classic ques- tions of what and how much to pro- duce, and who gets what products.''4 Cost- benefit analysis is a subfield of eco- nomic analysis: a specific application of economic analysis to nonmarket activity. We have defined cost-benefit analysis as the analytical examination of the costs and benefits of alternatives designed to meet specified objectives under various contingencies or states of the world. 5 Some differentiate cost-benefit analysis from cost-effectiveness analysis; the for- mer referring to long-range financial ef- fects (e.g., increased dollar income) and the latter to short-range measured output in nonfinancial terms (e.g., num- ber of books circulated) .6 Although sys- tems analysis has been used to refer to cost-benefit analyses, because its use is much more widespread, having applica- tion in areas from computer technology to political analysis, we define systems analysis as the study of systems or com- plexes or organized and interrelated parts, in terms of inputs, outputs, and internal functioning. 7 Our definition of cost-benefit analy- sis has already included most of the elements of the basic analytical frame- work: costs, benefits, alternatives, and contingencies. What then is cost-benefit analysis? It is a way of looking at the world. Usually one starts from a set of objectives that a decision maker has in mind. The analyst finds measures of the Library Decision Making I 413 extent to which the objectives may be met. For example, if an objective of a library were to provide reading material to library users, then one measure of meeting this objective would be annual book circulation. One then examines the alternatives for fulfilling each objec- tive. By constructing models (e.g., formu- las, computer simulations), the analyst relates each alternative to its correspond- ing costs and benefits (i.e., the degree to which objectives are met). The model is used across several contingencies or states of the world. Given the costs and benefits associated with each alternative in each contingency, a criterion or mea- sure of preferredness (e.g., maximizing profits) is selected and the "best" al- ternative is chosen. Figure 1 illustrates the cost-benefit procedure. Note that the method is actu- ally circular-objectives are revised in light of feasibility and costs, new alter- natives are created, models are refined, and the decision process is continuously in motion. Brief Critique The elements of the cost-benefit analysis model serve as the basis for a brief critique of the method. Attempts to define library objectives can lead to clarifications of purpose, yet they often result in futile searches for well-hidden goals obscuring the true clients of the library. 8 While efforts to generate al- ternatives to perform library services more efficiently and effectively are made, the question of the practicality and fea- sibility of radically different ideas weights the ultimate analysis against in- novative options. Relating costs to al- ternatives becomes the key task, and numbers generated through cost model- ing become the foci of economy drives. Benefit modeling, however, is weakest when the alternatives are most innova- tive, e.g., public library programs based 414 I College & Research Libraries • November 1974 r:::J ~ \ \ Fig. 1 The Basic Cost-Benefit Framework on distributing paperback books for dis- advantaged patrons, university libraries handing out free copies of required ar- ticles to students. 9 Because the contin- gencies studied most are those thought · to affect costs, not benefits, inflation of prices is emphasized over inflation of goals in serving users. Each of these difficulties is related to the political context in which the eco- nomic study is conceived, implemented, and received. The basic political prob- lem with economic analysis transcends operational and day-to-day difficulties and political intrigue. The basic politi- cal problem centers on political conflict inherent in all our institutions, includ- ing libraries. It is this conflict that is in- appropriately dealt with or ignored in economic analysi~. The Political Problem with Economic Analysis: An Example Near the end of the data collection stage, the economic systems .analysis of the Massachusetts Institute of Technol- ogy ( M.I.T.) libraries yielded a list of twenty library alternatives with associat- ed costs and benefits.1o At this stage in the analysis several points were evident: 1. Several alternatives were not worth considering. Alternatives offering fewer benefits at a higher cost than comparable alternatives were dis- carded. For example, storing books off campus rather than on campus saved no money and resulted in losses of benefits to library users. 2. Many conclusions with major pol- · icy implications were already war- ranted. For example, inexpensive storage appeared to offer little fi- nancial savings at a fairly high cost in benefits to a majority of the M.I.T. community. 3. Although alternatives could be de- scribed in terms of costs and bene- fits with respect to the two major library objectives, more informa- tion was required to select and rec- ommend a subset of alternatives. For example, should the cost of reproducing copies of library ma- terials be . reduced or should many course-required articles be distrib- uted free of charge~ Should either be done with or without a decen- tralization of library space?11 It is at this point that cost-benefit analysis (and economic analysis) comes to a grinding halt. Cost-benefit analysis assumes that the objectives, even if un- clear at the beginning of the analysis, can be specified at some point to the satisfaction of the decision maker. Cost- benefit analysts recognize that multiple objectives may exist and suggest that the tradeoffs, the extent to which meeting one objective leads to a failure to meet other objectives, be specified and clearly displayed. But did our analysis indicate objectives that could be agreed upon? We decided to present the data on al- ternatives derived from our analysis to the individual members of the univer- sity community, thus to allow each to act as if he or she were the ultimate de- cision maker.12 Because it would have been too costly to reach all members of the community, we drew a random sam- ple of undergraduate students, graduate students, and faculty and research staff and presented them with twenty alterna- tive changes, with a brief description of costs .and relevant benefit considerations, for the M.I.T. libraries. Respondents were given budgets of $0, $100,000, and $200,000 to spend for changes in the li- braries. The analysis of the survey clearly in- dicates that different subgroups of the M.I.T. community either had different objectives in mind or viewed different means as being best for meeting com- mon objectives: The general conclusion is that the three major campus groups differ in the systems they would like the library to adopt. Undergraduates seek to ex- pand and centralize the reserve collec- tion by cutting research services. Grad- uate students add lower Xerox prices and increased access to this list of de- sired systems and would prefer to cut seating rather than cataloging. The faculty are the most willing to alter Library Decision Making I 415 book storage and cataloging and rela- tivefy less desirous of a centralized re- serve system .... The less a respondent reported using the libraries the more likely he was to select saving money on book storage and seating and to spend it on lower Xerox rates, departmental li- braries, and an all-Xerox reserve sys- tem. Low users thus tend to be out- side-use oriented. The high users pre- fer expanding seating, acquisitions, reference, and access to other collec- tions. The high users thus are research oriented. We have concluded that the library has traditionally served one cli- entele, the research oriented. There now appears to be, however, a second clientele, who spend few hours in the library and seek not the space but the materials in its collection. We believe, with as yet no proof, that many of those oriented to outside use prefer to work outside the library but are forced, primarily by the reserve system, to work in the library. We hypothesize that these users (and many other po- tential users) could be served by a li- brary emphasizing distribution as well as in-house facilities and services.13 The M.I.T. analysis indicates that the alternatives faced by the M.I.T. library and university administrators involved major choices among various subgroups on campus. Furthermore, the analysis strongly suggests that decisions now fa- vor faculty far more than students. The political problem with economic analysis is that there is no economic way to resolve differences among al- ternatives meeting different objectives held by different subgroups; where po- litical conflict exists a political solution must be found. 14 This is not news to most economists. What library decision makers require is help in resolving these political conflicts. Presumably political analysis can help. POLITICAL ANALYSIS Political scientists would not agree on the nature of analysis necessary to deal 416 I College & Research Libraries • November 1974 with such political conflict. Some would argue that an analysis should begin with a positivist or descriptive analysis of li- braries with a focus on who decides and by what process. Others would argue that an explicitly normative or value based analysis, with a major focus on issues of equity and responsiveness, is required. Because neither of these ap- proaches has been applied' to libraries, a first step falling between the norma- tive and positive poles of political sys- tems analysis has been chosen here. Be- low, David Easton's descriptive frame- work is used to raise the normative ques- tions which library decision makers should be addressing.15 Easton defines politics as the authori- tative allocation of values for a society. In the past, many governmental institu- tions, perhaps education is the best ex- ample, have been viewed as being out- side of the realm of politics. In 1969, in an introduction to a reader on the politics of education, the editor stated that "The idea that politics and public education are intimately related was practically unthinkable as recently as a decade ago. . . . At the very lea~t, any governmental process involving authori- tative decisions on matters of public rel- evance is of a political nature."16 Thus an entire literature dealing with the pol- itics of education has developed.17 Cer- tainly it would not be inappropriate to raise issues concerning the politics of public libraries and libraries at public universities. Studying the politics of university li- braries derives from another expansion of political analysis to the area of pri- vate government. 18 Public governments have been defined as "those general as well as special-purpose associations and agencies either to which all inhabitants of a given locality are subject or of which all citizens are members"; and private governments are "those limited- purpose associations or organizations, usually voluntary in membership, which exist both alongside and subordinate to public governments."19 Examples of private governments are corporations, trade unions, professional associations, and universities. Indeed, the public ver- sus private distinction has become in- creasingly blurred, especially as applied to universities, within the past decade.2o The basic questions one asks about pri- vate governments are political: Are (and can) private governments (be) democratic?21 Related questions include: Who gets what, when, and how?22 Although many alternative models of the political process exist, I believe that Easton's framework provides a useful analytical scheme for beginning a po- litical analysis of libraries.2a Easton's Framework for Political Analysis Easton's model (see Figure 2) is sim- ple in its conception but complex in its full description. Dye describes the the- oretical framework succinctly: One way to conceive of public policy is to think of it as a response of a po- litical system to forces brought to bear upon it from the environment. Forces generated in the environment which affect the political system are viewed as inputs. The environment is any con- dition or circumstance defined as exter- nal to the boundaries of the political ·system. The political system is that group of interrelated structures and processes which functions authorita- tively to allocate values for a society. Outputs of the political system are au- thoritative value allocations of the sys- tem, and these allocations constitute public policy. Systems theory portrays public pol- icy as an output of the political system. The concept of "system" implies an identifiable set of institutions and ac- tivities in society that function to transform demands into authoritative decisions requiring the support of so- ciety. The concept of "systems" also l. .. Library Decision Making I 417 0 I Demands U N THE T Decisions and p U p POLITICAL U T Support SYSTEM Actions T s s ~-------____..._) Fig. 2 The Systems Model implies that elements of the system are interrelated, that the system can respond to forces in its environment, and that it will do so in order to pre- serve itself. Inputs are received into the political system in the form of both demands and support. Demands occur when individuals or groups, in re- sponse to real or perceived environ- mental conditions, act to affect public policy. Support is rendered when in- dividuals or groups accept the out- come of elections, obey the laws, pay their taxes, and generally conform to policy decisions. Any system absorbs a variety of demands, some of which conflict with each other. In order to transform these demands into outputs (public policies) , it must arrange set- tlements and enforce these settlements upon the parties concerned. It is rec- ognized that outputs (public policies) may have a modifying effect on the en- vironment and the demands arising from it, and may also have an effect upon the character of the political sys- tem. The system preserves itself by: ( 1) producing reasonably satisfying outputs, ( 2) relying upon deeply root- ed attachments to the system itself, and ( 3) using, or threatening to use force.2 4 POLITICAL ANALYSIS AND UNIVERSITY LIBRARIES System Boundaries and Legitimacy The first question that arises is wheth- er we can determine the boundaries of a political system. Throughout the M.I.T. library study we felt too con- strained by the definition of the system we were studying, "the M.I.T. libraries." The use and evaluation of a university library are not independent of the book stores within (and without) the univer- sity. To declare one a legitimate item for analysis and the other as outside of the area of analysis may be to miss the dynamics of the situation. It was sur- prising to discover that a high-level, li- brary acquisitions department staff member had not only made no effort to buy books from the Harvard Coop but also had never even been to this store, one of the world's largest bookstores. We were surprised to receive veiled threats by a department chairman after 418 I College & Research Libraries • November 1974 we had measured his departmental li- brary's floor-space without his permis- sion. The quality of departmental li- braries must surely determine the nature and degree of use of the main libraries. What units should be included in the library decision maker's domain? Output and Benefits Unlike Easton, it seems that the most fruitful political analysis must begin with the output stage of the political process. The analysis of output, done within the cost-benefit framework of the M.I.T. study, provides some signifi- cant information and raises some im- portant questions. The best tool available for analyzing policy was ( and often is) the library budget. But budgets are usually input based (e.g., cost of books purchased, cost of personnel salaries) rather than output or policy derived (e.g., cost of providing student services for course- work). Perhaps even more interesting, the M.I.T. library budget, divided into parts among discipline-related libraries, was considered confidential. To para- phrase one 'library administrator, "If the social scientists knew what we were spending on the physical science library, they'd start asking for more funds." A program budget analysis of the M.I.T. library seems to show quite clear- ly who benefits from the current deci- sion-making system. Only 23 percent of the total budget is used for providing required reading and facilities for studying, i.e., 'less than a quarter of the budget is devoted to nonresearch, course- related studerit services. Of course, this overstates the antistudent bias, for un- dergraduate and particularly graduate students devote much effort to research both inside and outside of courses. As noted in the discussion of the cost-benefit analysis of the M.I.T. li- braries, the survey analysis challenges the myth of a unitary community.25 In fact it suggests that the allocation of benefits, if not costs, is weighted in fa- vor of faculty and staff. Why should this be so? Demand and Democratic Process The concept of demand is a crucial one in Easton's framework. An analysis of demands made upon library adminis- trators at M.I.T. would probably indi- cate that a small number of senior fac- ulty are the primary demanders. The li- brary advisory committee contained no students; the administrators themselves spoke almost entirely of faculty com- plaints.26 What channels, both formal and informal, are required for those affected by decisions concerning libraries to be adequately heard? The concept of demands is too limit- ed for the political analysis of quasi- public institutions like university li- braries. Few preferences, defined as de- sired states of affairs, even reach the level of demands. Easton concentrates on the reasons for the weeding of de- mands and the attrition of preferences in the input stage of the political pro- cess. We should ask, as does Easton, what institutions exist to filter and chan- nel demands to library decision makers? How successful are different kinds of people within the university community in making their demands heard? To what extend should access be equalized? The mobilization of bias should also be considered crucial by library analysts. Several political scientists have criticized their discipline for the substantive con- clusion that American institutions are open and responsive to minority groups.27 They argue that this optimistic substantive conclusion derives in part from a methodological problem, analyz- ing only decisions made by public bodies. Backrach and Baratz ask, "Can the researcher overlook the chance that some person or association could limit decision-making to relatively noncontro- versial matters, by influencing communi- ty values and political procedures and I I ·~ ... rituals?"28 By limiting political analysis to overt decisions, the role that elites play in mobilizing bias, i.e., in defining the nature and states of the political game, is overlooked. The mobilization of bias plays a criti- cal role in library policy. One of our early suggestions at M.I.T. was that the price of reproducing pages of library materials within M.I.T.'s libraries should be reduced. Although the price was lat- er decreased, the action was based upon an agreement that decreasing the price would ultimately increase revenue (i.e., elastic demand) and the system would remain self-supporting. But why should the dissemination of information by copying be self-supporting, and who is disadvantaged by this decision rule? Whereas many faculty have research grants, departmental resources, and rela- tively high incomes, students are at a rel- ative disadvantage in the marketplace. Libraries do not break even on provid- ing books. Shoul~ they break even on copying materials for dissemination? One economist has made an argument that could have been based on the mo- bilization of bias concept: Or why do not librarians diminish their stock of hard-cover books and acquire in their stead substantial in- ventories of paperbacks which they would then give away free? We are inclined to reply, "Why, that would be crazy: our budget would soon be exhausted." And yet that is exactly what librarians are doing now except instead of giving books away free they are giving staff services away free.29 Keller calls for implicit (or explicit) pricing of library services. 30 Easton's framework, indeed all po- litical frameworks, should include a ba- sic economic concept of exit. Hirsch- man argues that one mechanism of voic- ing disapproval within the political as well as economic sector is exit, e.g., leav- ing the organization or not consuming the product. 31 To what extent do po- Library Dec.ision Making I 419 tential library users seek other sources of information because of library in- effectiveness? To what extent do some groups lack an effective means of in- fluencing library decision makers by their inability to exit? The larger question that each of these points concerning demand raises is the appropriateness of democratic norms fo·r library decision making. Should libraries be run more democrati- cally than they now are? Decision Making and Selecting Decision Makers The analysis of decisions and deci- sion makers is a crucial aspect of Eas- ton's framework The analysis of li- brary decision making must reach be- yond the traditional organizational bounds of the exercise of rationality. Lakoff has expressed the criticisms of traditional organizational analysis as follows: The study of organizational decision- making studiously avoids asking the kind · of questions that would render the · study of decision-making genuine- ly political. It does not ask what con- stituencies are involved, or how the legislative is related to the executive, or how the authority of the decision- maker is made accountable to those he represents. It does not ask whether the system is constitutional or just, legitimate or illegitimate. Instead the study of decision-making in organiza- tions is confined to the question of whether and to what extent the func- tions of management are exercised ra- tion~.lly. The stress, in other words, is clearly on administration rather than government, on the integrative func- 'tion of social organization, on improv- ing the efficiency of the decision-mak- er. There is pnictically no attention paid to the question of whether people who are members of the ox:ganization or who are served by it have or ought to have control over it, whether they have any right (a term which would probably be considered altogether un- 420 I College & Research Libraries • November 197 4 scientific by students of organization) to be consulted in the decision-making process or indeed to decide what form the process will take. 32 In this conventional sense, the study of organizations, despite its focus on decision making, has been quite apoliti- cal. Dye's identification of barriers limit- ing rational decision making, in many ways analogous to criticisms of the use of cost-benefit analysis in libraries, might serve as a starting point for an analysis of library decision making. They are restated below as hypotheses. 1. There are no community values which are usually agreed upon, but only the values of specific groups and individuals, many of which are conflicting. 2. The many conflicting values can- not be compared or weighted: for example, it is impossible to com- pare or weight the value of indi- vidual dignity against the loss of rare books. 3. The environment of library poli- cy makers, particularly the power and influence system, renders it impossible for them to see or ac- curately weight many community values, particularly those values which have no active or powerful proponents. 4. Library policy makers are not mo- tivated to make decisions on the basis of community goals, but in- stead try to maximize their own rewards-power, status, money, etc. 5. Library policy makers are not mo- tivated to maximize net goal achievement, but merely to satis- fy demands for progress; they do not search until they find "the one best way" but halt their search when they find an alternative which "will work." 6. Large investments in existing pro- grams and policies (e.g., catalog- ing systems, library buildings, and other "sunk costs") prevent policy makers from reconsidering alter- natives foreclosed by previous de- cisions. 7. There are innumerable barriers to collecting all of the information required to know all possible pol- icy alternatives and the conse- quences of each alternative, in- cluding the cost of information gathering, the availability of the information, and the time in- volved in its collection. 8. Neither the predictive capacities of the social and behavioral sci- ences nor the predictive capacities of the physical and biological sci- iences are sufficiently advanced to enable policy makers to under- stand the full range of conse- quences of each library policy al- ternative. 9. Library policy makers, even with the most advanced computerized analytical techniques, do not have sufficient intelligence to calculate accurately cost-benefit ratios when a large number of diverse politi- cal, social, economic, and cultural values are at stake. 10. Library policy makers have per- sonal needs, inhibitions, and in- adequacies which prevent them from performing in a highly ra- tional manner. 11. . Uncertainty about the conse- quences of various policy alterna- tives compels policy makers to stick as closely as possible to pre- vious policies to reduce the likeli- hood of disturbing, unanticipated consequences. 12. The segmentalized nature of pol- icy making in large library bu- reaucracies makes it difficult to co- ordinate decision making so that the input of all of the various specialists is brought to bear at the point of decision. 33 '· Testing these hypotheses requires an analysis of the values and personal goals of library decision makers, the power of competing interests in universities, the incentive structures surrounding li- brary administrators, and the nature of information available to decision mak- ers. Significant normative questions fol- low. Should library decision makers be more representative of those who use the libraries? Should the incentive struc- ture for advancement in library admin- istration be altered to better reflect user and potential user demand? LESsoNs FOR LmRARIANs Unfortunately, whereas economists can advocate that library decision mak- ers try to maximize benefits at a given budget level, political scientists can of- fer no clear-cut decision rule as an al- ternative. This paper ends with ques- tions that library decision makers should ask, but no simple answers of what ac- tions to take if answers are found can be offered at this point. 34 Future re- search is needed, although this obviously will not solve all the political problems of librarians. Following the more complete Easton model, library decision makers should ask themselves: 1. What is the relevant library sys- tem? Have I excluded a key com- ponent that determines user behav- ior but has traditionally fallen out- side my purview? Can I coordinate decisions between my area and the additional area? 2. What are the environmental con- straints that appear to limit my discretion? Can they be altered? 3. What groups (and individuals) make demands? Are they represent- Library Decision Making I 421 ative of the potential users of the library? What preferences do not become demands? Are new or re- vised mechanisms needed to en- courage more demands? 4. What is the general climate of opinion with respect to the library, e.g., support for library? Has the climate provided me with so much latitude that demands remain un- met? What assumptions (of users or administrators) limit the con- sideration of alternative policies? Who benefits from these assump- tions? Who does not benefit? Can the asumptions be changed? 5. Who plays a role in decisions about library allocations? To what extent are users or potential users in- volved? To what extent are those affected by decisions helping to make them? 6. Who benefits from (and pays for) the library? Does the budget show this? What services serve what groups? How well are they served? 7. What feedback is available to the decision maker to evaluate current allocations? What mechanisms for feedback exist? Are they successful in bringing evaluations of users to decision makers? Do nonusers have access and do they use feedback systems? In brief, political systems analysis is analogous to economic systems analysis: it is a way of thinking. Woodrow Wilson, asked whether he had much difficulty in accustoming him- self to practical politics, stated that af- ter his experience in university politics at Princeton everything else seemed sim- ple. It is time that we all recognized the politics of libraries and acted according- ly. REFERENCES 1. Jeffrey A. Raffel and Robert Shishko, Sys- te-matic Analysis of University Libraries: An Application of Cost-Benefit Analysis to the M.I.T. Libraries (Cambridge, Mass.: 422 I College & Research Libraries • November 1974 M.I.T. Press, 1969). Our M.I.T. editor said goodbye to us with the comment, "I hope you get some reviews." When we replied, "You mean some good reviews," she re- sponded, .. No, just some reviews." 2. General references in the area of cost-bene- fit analysis include Charles I. Hitch and Roland N. McKean, The Economics of De- fense in the Nuclear Age (Cambridge, Mass.: Harvard Univ. Pr., 1960); and Da- vid Novick, ed., Program Budgeting: Pro- gram Analysis and the Federal Budget (Cambridge, Mass.: Harvard Univ. Pr., 1965). References on the .. economics of in- formation" can be found in Harold Anker Olsen, The Economics of Information: Bibliography and Commentary on the Lit- erature (Washington, D.C.: ERIC Clear- inghouse on Library and Information Sci- ences, 1971). (ED 044 545). 3. Mancur Olson, .. Economics, Sociology, and the Best of All Possible Worlds," The Pub- lic Interest, no. 12 (Summer 1968), p.96- 118. 4. Olsen, Economics of Information, p.l. 5. Jeffrey A. Raffel and Robert Shishko, .. Cost- Benefit Analysis for Library Administra- tors," paper presented to Massachusetts Chapter of the Special Libraries Associa- tion on March 11, 1969, in Boston, Massa- chusetts. 6. Melvin R. Levin and Alan Shank refer to cost-benefit analysis "as a measurement technique in which the total costs of a giv- en project or program are compared with the probable total benefit .... The result is a numerical ratio .... " Cost-effectiveness is viewed as a variation of cost-benefit analy- sis where the output is expressed in "raw form without conversion to dollars." See Levin and Shank, eds., Educational Invest- ment in an Urban Society: Costs, Benefits, and Public Policy (New York: Teachers College Press, 1970), p.1-2. Using this distinction, I refer primarily to cost-effectiveness analysis in its applica- tion to libraries and cost-benefit analysis as the general approach. 7. Raffel and Shishko, "Cost-Benefit Analysis," p.2. 8. Late in the M.I.T. analysis it became evi- dent that an advertising objective of the li- brary was also significant; that is, librarians wanted to encourage those in the univer- sity community to increase their use of the library. Thus attempts were made to en- tice book use by lavishly furnishing lounges in the libraries. This goal was rarely ac- knowledged. 9. In the M.I.T. study we tried to resolve this problem by asking potential users to judge benefits themselves. 10. Raffel and Shishko, Systematic Analysis of University Libraries, p.50-55. 11. Ibid., p.46-67. 12. Terry N. Clark, in an unpublished paper titled "Please Cut the Budget Pie" ( re- search paper # 37 of the Comparative Study of Community Decision-Making, Summer [1972]), develops a further ration- ale and somewhat different methodology to measure citizen preferences for various public policies. 13. Raffel and Shishko, Systematic Analysis of University Libraries, p.65. 14. There is a literature in economics (welfare economics and public choice economics) on this subject. The proposed solutions in- clude: ( 1 ) maximizing total utility across individuals-but this requires the interper- sonal comparison of utility and the measure of utility, both problematic procedures; ( 2) transforming costs and benefits into dollars and maximizing the net figure-but the full transformation is usually impossible and questions like those raised in the first method still arise; and ( 3) .only taking those actions that make no one worse off and at least one person better off-but this case arises infrequently in the era of de- clining or steady budgets. I am indebted to the critiques of cost- benefit analysis by Aaron Wildavsky. See his .. The Political Economy of Efficiency: Cost-Benefit Analysis, Systems Analysis and Program Budgeting," Public Administration Review 26:292-310 (Dec. 1966); and "Rescuing Policy Analysis from PPBS," in Public Administration Review: PPBS Re- examined 29:189-202 (March/ April1969). 15. David Easton, A Systems Analysis of Po- litical Life (New York: Wiley, 1965). 16. Alan Rosenthal, ed., Governing Education: A Reader on Politics, Power, and Public School Policy (Garden City, N.Y.: Anchor Books, 1969 ), p.viii-ix. 17. See Frederick Wirt and Michael Kirst, The Political Web of American Schools ( Bos- ton: Little, 1972), for one volume apply- ing Easton's framework to the politics of education. 18. See Sanford A. Lakoff and Daniel Rich, eds., Private Govermnent: Introductory Readings (Glenview, Ill.: Scott, Foresman, 1973), for an excellent discussion of the rationale for studying the politics of private government and case studies, including the politics of university governance. 19. Ibid., p.3. 20. See Sanford Lakoff, .. Private Government in the Managed Society," in Lakoff and Rich, eds., Private Government, p.218-42. 21. Lakoff and Rich, Private Government, preface. 22. Harold Lasswell, Politics: Who Gets What, When, and How ( Glencoe, Ill.: Free Press, 1958). 23. Thomas Dye recently described several models used in analyzing public policy: (a) elite-mass model, (b) group model, ( c ) incremental model, (d) institutional model, (e) systems model, and (f) ration- al model. See Thomas R. Dye, Understand- ing Public Policy (Englewood Cliffs, N.J.: Prentice-Hall, 1972). 24. Dye, Understanding Public Policy, p.18- 19. Dye's conceptualization is based upon David Easton, "An Approach to the Analy- sis of Political Systems," World Politics 9:383-400 (1957); and Easton, A Frame- work for Political Analysis (Englewood Cliffs, N.J.: Prentice-Hall, 1965). 25. In the preference survey at the $0 budget level, although differences were small, fac- ulty were somewhat more satisfied as mea- sured by the percentage desiring changes from current allocations. 26. The reader should note that this paper is not an indictment of M.I.T. in particular, although there is evidence that the libraries are governed in the same way as are other elements of the university (e.g., health ser- vices, graduate school). 27. These arguments relate to a larger battle among elitists, pluralists, neoelitists, and so on. See Dye, Understanding Public Pol- icy. 28. Peter Bachrach and Morton S. Baratz, Power and Poverty: Theory and Practice (New York: Oxford Univ. Pr., 1970), p.9. 29. John E. Keller, "Program Budgeting and Cost Benefit Analysis in Libraries," College & Research Libraries 30: 160 ( March 1969 ) . 30. It should be noted that many university li- braries charge fees for use, for reproduction and for organizational users. My own opin- ion is that fees for the former are far too high and the latter far too low. In any event, both require further political analy- sis. 31. Albert Hirschman, Exit, Voice, and Loyal- Library Decision Making I 423 ty (Cambridge, Mass.: Harvard · Univ. Pr., 1970). . 32. Lakoff, "Private Government in the Man- aged Society," p.229. 33. Dye, Understanding Public Policy. 34. Note that analogous questions arise about public libraries. Political scientist Edward Banfield is one of a few people to raise ex- plicit political questions about public li- braries. Banfield begins his discussion of urban libraries with the question of their purpose. "It [the urban library] is trying to do some things that it probably cannot do, and it is doing others that it should not do." In Easton's terms the question expands: ( a) What is the relevant system? Educa- tional institutions? Information institutions? (b) What are the outputs of the library? What groups does it serve? How well does it serve them? (c) Why does the library try to serve these. groups? What denl:ands does it try to meet? What preferences nev- er become demands? What tasks might it accomplish that are not now viewed as ap- propriate? Banfield argues that libraries should serve "serious," not "light," readers; the lat- ter group could be served by rental and paperback libraries. But most serious read- ers can pay for the services they receive. Should the general public then pay for a subgroup to receive the services? Perhaps what Banfield is trying to com- municate is that although the cost of li- brary services may serve as the focal point for library decision making and public con- cern (e.g., closing and reduction in services of public libraries), the basic problem of public libraries is not an economic prob- lem. Rather, we have not examined library priorities and reallocated library resources to meet changing political circumstances. Only a political analysis will indicate why this is so and what changes should be made. See Edward C. Banfield, "Some Alterna- tive~ for the Public Library," in Ralph W. Conant, ed., The Public Library and the City ( Cambridg~, Mass.: M.I.T. Press, 1965), p.l02-13.