ACRL News Issue (B) of College & Research Libraries 5 0 4 / C&RL News Librarian salaries: Paid what we’re worth? By Richard W. Meyer Associate D irector Clemson University Library Librarians have made some progress in the market place. H erb W hite expressed th e concern o f many librarians w hen he posed th e question, “Why don’t we get paid m ore?”1 Although th e issue of earnings may be fraught w ith em otion, this question takes on m eaning only w hen a fram e o f reference is added: “m ore com pared to w hom ?” In the case o f academ ic librarians, th e com parison is relevant to teaching faculty; th a t is, to those people at colleges and universities w ho get paid to teach students and to work on favored research projects. Librarians often perceive them selves to be u n d e r­ paid relevant to this group. Those perceptions are usually based on a com parison o f anecdotal evi­ dence or on average salary data. O ften, th ese comparisons look at faculty and librarians in similar ranks.2 W hen d ata is available, librarians as a group always seem to b e lu m p ed in w ith th e instructors at average salaries substantially below even assistant professors. C om parisons seldom include any infor­ mation on how salaries are d ete rm in e d in general, nor do they include com parisons o f large groups o f 1H erb W hite, “W hy D o n ’t W e G et Paid M ore?” Library Journal 111 (M arch 1, 1986): 70-71. 2John B uschm an, “T h e F lip Side o f F acu lty S ta tu s,” C & R L N e w s 50 ( D e c e m b e r 1989): 972-76. data on individuals. T he in ten t here is to provide a b e tte r fram ework for understanding by exploring th e factors related to determ ination o f salaries in th e m arket and by sum m arizing an examination o f salaries at th e individual level. Some com parisons o f salaries have b een m ade using statistical techniques to examine data on individuals such as teaching faculty and adm inistra­ to rs.3 Additional work has exam ined th e various d e te rm in a n ts o f librarian salaries from overall viewpoints.4 These studies often focus on th e issue o f sex as a d eterm inant o f salary and they appear to 3D avid A. Katz, “Faculty Salaries, Prom otion, and Productivity at a Large University,” American Econom ic Review 63 (June 1973): 469-77; Bharat R. Kolluri and M ichael J. P iettee, “D eterm inants of th e Salaries o f C h ief Academic A dm inistrators,” A tlantic Economic Journal 13 (July 1985): 61-89. 4A nita R. Schiller, Jam es K. Grim m , and Margo T ru m p e te r, Characteristics o f Professional Person­ nel in College and University Libraries (Spring- field, Ill.: Illinois State Library, 1969); Ung C hong Kim, A Statistical Study o f Factors Affecting Sala­ ries o f A cadem ic L ibrarians at M ed iu m -S ized State-Supported Universities in Five M idwestern States (P h.D . dissertation, In d ian a U niversity, 1980). June 1990 /5 0 5 show that, other conditions held constant, male librarians make higher salaries than women.5 F u r­ ther study by others indicate that the library labor market is substantially free of control by employ­ ers.6 The erection of barriers to restrict the supply of librarians and bargaining to increase earnings among academic librarians appear to be issues unexplored by systematic means. Given this envi­ ronment, a general overview of market factors provides illumination of the causes that establish earnings levels. This overview can then be followed with a description of how librarians have attem pted to influence market forces and what success they may have had. Salaries of librarians are determ ined in the market place, just like salaries in general or the price of any commodity. However, librarians ap­ pear to exert some force in the market that affects their salaries positively. W hen people perceive their earnings to be low, an effort may be made to do something about it. In the case of academic librarians that effort may have been underway for some time, but remain largely unrecognized for lack of understanding of the economic forces at work. The rest of this paper proceeds in three sections. First, an overview is provided of the general forces driving wage rates. Second, forces that librarians have already exerted to improve earnings are d e­ scribed next. Third, a report on an analysis is sum­ marized which indicates that academic librarians may be paid salaries comparable to faculty. A complete report of the econometric analysis will appear in a paper to be published later. Fourth, some additional factors are suggested that may affect librarian salaries. These factors need to be addressed if librarians expect to achieve still higher earnings. Some final thoughts conclude the paper. The factors that establish wages The market determines salaries in the absence of restrictions to open competition? At any instant in time, at a given wage, there will be an equilibrium between supply and dem and that will determ ine the num ber of laborers employed. Salaries are determ ined by the interaction of the supply of human capital available through laborers and the 5David R. Dowell, “Sex and Salary in a Female Dominated Profession,” Journal o f Academic Li- brarianship 4 (May 1988): 92-98. 6Judith S. Braunnagel, “Job Mobility of Men and Women Librarians and How It Affects C areer Advancement,” American Libraries 10 (D ecem ber 1979): 643. 7Lloyd G. Reynolds, Stanley H. Masters, and Colletta H. Hoser, Economics o f Labor (Engle­ wood Cliffs, N.J.: Prentice-Hall, 1987). dem and for that capital by employers. Competition may be restricted by monopoly effects such as collusion through trade unions or other organiza­ tional approaches. Market size or labor market immobility may also limit competition by providing too few jobs to allow individuals promotional op­ portunity. Models existing in the economic litera­ ture to explain salary levels generally focus on either the dem and side where conditions are con­ trolled by the employer or the supply side, which depends on the characteristics of individuals such as hum an capital appreciation. These are not com peting models, because both dem and and supply considerations need to be understood to get a clear picture of how the market determ ines sala­ ries. One basic economic model shows that, from the employer’s point of view, the salary offered to an individual is driven by the equilibrium of marginal revenue product (MRP) and salaries dem anded by labor. Firms will hire labor up to the point where the cost of the next unit of labor just equals the marginal revenue product of that unit. Marginal revenue product is the added revenue to the firm from the last unit produced. Firms will stop pro­ ducing and thus stop hiring labor, at the point where the cost of producing the last unit just equals the revenue from that unit. MRP may does not remain constant for a variety of reasons. MRP may increase at first, because four hands can often be more efficient than two. For example, two people can hang a piece of paneling easier than one person. A third may make the project of paneling even easier, but after a fourth or fifth laborer is added, efficiency begins to decline unless labor is assigned to different tasks. D epend­ ing on the industry and the wage rate, the optimal amount of labor will occur at different levels. As a library example, consider the num ber of librarians working at a reference desk at the same time. After the second or third reference librarian is scheduled at the desk, they start getting in each others way. As a result, while two librarians may be able to handle more than twice as many students as one librarian, it doesn’t hold that you can add librarians to the schedule indefinitely and continue to work effi­ ciently. Since universities are engaged in non-profit ac­ tivities, it is less clear how they determ ine the level of salaries to offer prospective employees. How­ ever, it is likely that they have downwardly sloping dem and curves for labor. Instead of an MRP curve, non-profits have a measure of the extra value to constituents resulting from the added output of employees. Again, consider the reference desk. At some point, perhaps after the third librarian is scheduled at the desk, physical arrangements begin to make it difficult to serve patrons without stum­ bling over each other. The added value to users 5 0 6 / C & RL News from a third librarian may be very low com pared to the wage paid if the desk is too crowded for the librarian to serve users efficiently. W hen the cost of an additional unit o f labor exceeds the value of the added benefit to constituent users, that unit of labor will not be hired. Although it may be difficult to observe this as a conscious process, through the budget, the university communicates it’s decision on the num ber o f positions allocated to the library by means o f an evaluation based upon this kind of decision analysis. It may be easier to see how the university determ ines the level of salary appropri­ ate to each kind o f position and collection o f skills needed. Some collusion betw een campus administrators may be evident in that productivity analyses made by universities ten d to be limited to com paring average salaries for similar positions at similar insti­ tutions. Budget setting activities betw een the li­ brary and university administration related to sala­ ries at the institutional level usually involve nego­ tiations about comparisons of average salaries b e ­ tween institutions. However, as shown by others, this administrative collusion does not appear to have restricted mobility within the academic seg­ ment o f the profession.8 Therefore, the potential for imposition of em ployer control on librarian salaries may be limited to sharing o f salary inform a­ tion betw een institutions. However, this sharing of inform ation along with th e analysis described above, results in a determ ination o f the salaries the university is willing to pay for the particular job skills of individuals. Universities then ultimately make positions available with discreetly d e te r­ mined wages for specific attributes; so m uch for a reference librarian, so much for an adm inistrator, and so forth. W ho fills these positions depends upon those individuals who choose to develop specific hum an capital appropriate to the positions. To explain salary levels from the supply side, other models concentrate on employee based char­ acteristics such as choice and training. In accor­ dance with this hum an capital model, laborers increase their potential for higher salaries through acquisition o f education, on the job training, migra­ tion, and job search skills. The am ount o f schooling is particularly im portant to the individual as lever­ age for higher salary through increased productiv­ ity or as a sorting m echanism for employers select­ ing employees. On the job training, signaled to the employer through years o f experience, also plays a role in determ ining salary. In o ther words, an indi­ vidual’s salary is affected by the am ount of effort they extend to learn new skills through education and job experience. Given a higher level of skill, labor will dem and a higher level of salary. This is intuitively obvious; if you can’t get a higher level of 8Braunnagel, “Job Mobility.” wage for training acquired, why acquire it? F or example, many libraries are particularly concerned that reference librarians speak the same language as th e research users they serve. As a result they are willing to pay a prem ium for librarians who possess a second masters degree in an academic discipline. That, in turn, creates an incentive to librarians to acquire the hum an capital associated with a second degree. All o f this is not to say that either model determ ines salaries. Rather, it is the interaction of the factors that of these models represent, which establishes the salary of each set o f skills required for particular tasks that need to be accomplished. Librarians appear to be in the market at a p ar­ ticular equilibrium that reflects the value placed on th eir skills. W e can accept as given that there is no general monopsony effect existing; that is, the m arket is free of em ployer control of the wage rate. If wages available are too low in one location, librarians will decline to apply for jobs there. This will te n d to provide an incentive to employers to raise the wage rate, which ultimately increases interest in the positions. On the librarians’ side, individuals each decide how much effort to invest in the acquisition of skills. Ultimately, the average librarian develops a certain level o f education and experience, for which there is a certain maximum salary level that can be dem anded and still have reasonable success in finding work. At any point in time, an equilibrium occurs w here the total sm or­ gasbord of positions available equals the total col­ lection of skill packages offered by the collective group of librarians. From this point of equilibrium, some mechanism in addition to the hum an capital acquisition is required in order for librarians to increase earnings. This is w here collusion may come into play. Additional forces that may affect earnings Various attem pts at collusion have been used to positively affect the level of salaries of a group or profession. Generally, two approaches are avail­ able to the individual or group which is interested in increasing wealth. One approach is to seek addi­ tional wealth through increased production. The other is to seek to effect a transfer of wealth from another party or parties, such as employers. In the case o f labor, collusion offers a couple o f ways to accomplish a wealth transfer. One way to transfer wealth from employers to employees is to restrict the supply of laborers avail­ able. In a competitive market, a shortage of supply elevates the price that can be dem anded for a com modity or service. T he same is true of wages. F o r instance, the fewer doctors that are available to handle your medical problem s, the higher you likely are to bid for th eir skills. In order to restrict June 1 9 9 0 /5 0 7 supply, laborers have to create some barrier to entry into the trade or profession. Many trades accomplish this by requiring practitioners to be members of an organization or to possess specific attributes such as a certificate. Librarians appear to accomplish something similar to this by establish­ ing minimum credentials. Typically, the majority of advertisements for academic library positions carry the caveat that applicants must possess an MLS degree from an accredited institution and must dem onstrate academic library experience for non­ beginning librarians. Nancy Van House carefully explored the effects of market segmentation on the salaries of librarians.9 H er study shows that there is little mobility of librarians between special, school, public and academic segments of the profession even though their graduate school training is nearly identical. Therefore, entry into the marketplace as an academic librarian is largely restricted to begin­ ners. In economic terms, this is a barrier to entry, which restricts the supply of librarians. W hen sup­ ply is restricted, wage rates are elevated. A second way collusion may be effective is to bargain collectively for improved working condi­ tions, higher salaries, and so forth. On this basis, some gains may also be acquired by collusive ef­ forts through the profession to bargain with univer­ sity administrations for specialized status. The academic segment of the library profession has established the standard that librarians should be granted faculty status.10 Prelim inary results by others suggest that collusion exists in the form of wide-scale adoption of the standard requiring im plementation of faculty status among academic institutions. Recent surveys of the literature con­ firm that somewhere between 50 and 80 percent of academic libraries have adopted faculty status at least in part.11 Im plem entation of the concept may be an effort to collude for higher salaries from employers. If so, its effectiveness will be revealed in terms of a positive impact on salaries at those institutions with faculty status. The next section summarizes work that suggests that faculty status has been effective in elevating earnings of librari­ ans who have it. 9Nancy Van House, “Salary D eterm ination and O ccupational Segregation am ong L ibrarians,” Library Quarterly 56 (April 1986): 142-66. 10ACRL Academic Status Committee, “ACRL Standards for Faculty Status for College and Uni­ versity Librarians: A Draft Revision,” C&RL News 51 (May 1990): 402-404. 11Kee DeBoer and W endy Culotta, “The Aca­ demic Librarian and Faculty Status in the 1980s: A Survey of the Literature,” College & Research L i­ braries 48 (May 1987): 215-23; Emily W errell and Laura Sullivan, “Faculty Status for Academic Li­ brarians: A Review of the Literature,” College & Research Libraries 48 (March 1987): 95-103. Successful efforts to improve earnings A complete description of an econometric analy­ sis of three data samples related to librarian’s sala­ ries will be provided in a later article.12 A brief summary describing how librarians have collec­ tively attem pted to improve earnings will suffice to show that the academic segment of the profession has achieved some success. Some libraries have entered into collective bar­ gaining through a union to elevate the earnings, but this is not common. The cost of collective bargain­ ing involves substantial legal efforts along with the involvement of government officials who may be perceived as bureaucratic and hard to work with. A less expensive and more common approach to collective bargaining by librarians may be to enter into less formal negotiations with local university administrators on the basis of a professional stan­ dard. If the library staff at an individual university can convince the university administration that professional status for librarians is the accepted standard and would be good for the university, there may be little reason to withhold that status. Presumably acquisition of that status would also entitle librarians to salaries comparable to other faculty. That presum ption may be unwarranted, but the attem pt portends gains far in excess of the effort extended. At the request of librarians, some universities have endowed them with faculty status and some have not. Comparing salaries over a group of librar­ ies, roughly half of which have im plem ented fac­ ulty status, may reveal the success of this effort to improve earnings. The comparison must be made by looking at individual librarians in order to hold constant for differences in human capital apprecia­ tion. D ata was collected on all the librarians em ­ ployed at fourteen randomly selected institutions and at Clemson. Information collected on indi­ viduals included salary, race, sex, age, num ber of subordinates, num ber of publications, rank, tenure status, w hether the individual possessed a second masters and w hether employed in a main, law, business or medical library. This data was aug­ m ented with information on w hether the institu­ tion endows faculty status or not, the regional location, enrollment, w hether the institution is public or private, and the size of the city in which it is located. The salaries of all the individuals were regressed on the other variables using an ordinary least squares model. The regression analysis re ­ vealed that, holding constant for all the variables listed here, librarians earn approximately six p er­ cent higher salaries at institutions with faculty status. This is to say that, taking into account the l2Library Administration and Management, Fall 1990. 5 0 8 / C & RL News regional differences, enrollm ent, city size, and other institutional factors, as well as the major individual variables, librarians with faculty status earn more than those without. F u rth erm o re, a second analysis compared the salaries o f librarians at Clemson with about one- third o f th e teaching faculty selected randomly. Data collected included salary, race, sex, w hether the individual is an adm inistrator or not, educa­ tional level, publication rate, years at Clemson, and rank. A set o f dum m y variables to isolate the aca­ dem ic u n it in w hich each individual was employed was also included. Although this analysis was lim­ ited to C lem son, this university pays salaries that are similar to o th e r schools. In fact, the librarians at Clem son are indistinguishable from their peers at oth er universities in the first analysis. An ordinary least squares regression was run to determ ine the extent th a t these variables explain salary. T he analysis revealed that librarians at Clemson are paid salaries which are statistically indistinguish­ able from teaching faculty in the arts, humanities, social sciences, and forestry. They are paid less than faculty in sciences, engineering, business and architecture. These results are not inconsistent with subjective expectations. Scientists and engi­ neers with a Ph.D. can dem and and receive high salaries in the non-academic world, as can econo­ mists, finance experts, and so forth. Conversely, those in the hum anities, along with librarians, appear to have fewer opportunities in the non- academic world. The fundam ental point of the analysis is straight­ forward. Faculty status for librarians provides a low cost substitute for officially recognized collective bargaining through a union. F u rth erm o re, this substitute is nearly as effective as unions commonly are at raising wages. D epending on the industry, unions typically elevate wages by no more than about ten p e rc e n t.13 In addition, the analysis also indicates th at librarians can achieve equity with at least some segm ents of th e faculty, although there is little doubt that the average teaching faculty m em ber is b e tte r paid at Clemson. Regardless, librarians appear to have been somewhat success­ ful in doing som ething as a group about their perceived low pay by restricting entry to their segment o f th e profession. W hat’s more, those librarians who have been able to im plem ent faculty status are even b e tte r off. Added factors that have to be addressed One issue that may affect the supply-dem and equilibrium is the availability of substitutes for the services o f librarians. A lternative inform ation sources com pete with librarians. Anyone can use 13Reynolds, et al , Economics o f Labor. an open library. Although information seeking ac­ tivities o f a user may be less successful than if the assistance o f a librarian had been sought, self help may be less costly than the information transfer effort and em otional costs required to obtain effec­ tive help from a librarian. Even though librarians te n d to be very approachable at the reference desk, some cost to users is associated with securing assis­ tance. In addition to the effort needed to overcome natural reticence, it takes an effort to com municate o n e’s need sufficiently to get appropriate assis­ tance. This effort may be greater than that required to simply seek an answer oneself. To some people, it may even be less costly to do without the inform a­ tion than to bother a librarian. Anyone who has ever worked at a reference desk has experienced these characteristics of users. F urtherm ore, as inform a­ tion seeking skills improve, the relative cost of getting help com pared to doing it oneself, will increase. The num ber of faculty at Clemson who have their own online searching accounts with vendors, suggests that quite a few consider com ­ munication costs with a librarian to be higher than the effort required to learn how to help oneself. Two points that affect salaries also can be made about the relationship o f the supply dem and equi­ librium for librarian workers relative to the values of society and dem and for other kinds of work. First, at some point workers ten d to price th e m ­ selves out of the market, because of productivity limitations. Professional actors provide a simple example. Legitimate theatre has disappeared at a top professional level from all but a few cities in the U nited States. This is due to the limited audience available in the th eatre com pared to the movies. F or a given productive unit o f acting in the theatre, a small audience can be reached. That same p ro ­ ductive unit can reach far more people through a movie. However, most actors would set a similar price for an hour o f th eir services w hether they perform on stage or in a movie. O r at any rate, the opportunity cost o f stage work is lost movie work, which is often very highly paid. As a result, ticket prices at major theatres greatly exceed the price of movie tickets. O ne unit of viewing by a m em ber of an audience may be purchased for $5 or $6 at the movies. T hat same unit of viewing will cost $40 or $50 at the theatre. Therefore, dem and for theatre is greatly dim inished relative to movies. Given this low er dem and, many th eatres th ro u g h o u t th e country have closed and actors are out o f work. Similarly, over time, a higher and higher wage could be dem anded for the same unit o f librarian production. However, unless the rate of productiv­ ity o f librarians matches the increase in salary required, dem and for librarians will decrease. Or, conversely, librarians will not be able to require as high a real salary for th e same unit of production over tim e in order to maintain th e same level of June 1990 / 509 dem and for librarians. This phenom enon of declin­ ing real salaries may be mitigated by automation and other productivity enhancements. The use of integrated automation systems and utilities such as OCLC appears to have affected positively the pro­ ductivity of catalogers. Although it may be less easily observed, public access catalogs probably have increased the productivity of reference li­ brarians as well. In fact, online information re ­ trieval has certainly had this effect. However over­ all, these productivity increases have been less than sufficient to maintain real salaries of librarians at the same level over the past twenty years. Salaries of beginning librarians reported in 1967 were $7,305 p er year; in 1987 they were $20,874 per year. Librarians should have been at $23,971, just to stay even with the C onsum er Price Index.14 G reater productivity gains are required to maintain earnings, much less to boost them substantially. Second, the value society places on the produc­ tive output of particular kinds of labor seems to relate to the ease with which the consumer can express a description of the service provided. For example, physicians get a high rate of return for their labor. If you ask the average person what a doctor does, you will get some form of the same very terse answer: that is, a doctor cures people. On the other hand, if you ask what a librarian does, the answer will often be a vague expression of what a library is rather than what a librarian does. Intui­ tively, people can not be expected to pay a high price for something they can not even describe. Ergo, librarians’ salary demands may be limited by a lack of understanding of what it is librarians do. The point was made a few years ago by an adm in­ istrative dean who said that “academic administra­ tors and most faculty do not think about librarians; they think about libraries.”15 W hen librarians can make it more apparent what value they can pro­ vide, they may be able to dem and a higher price for their services. W ithout getting into a theory of value, it may be noted further that a hierarchy of values may be operable here. Consider the effect of illness on one’s lifestyle. Serious illness may make it ex­ trem ely difficult or even impossible to carry on any m anner of activity. Consequently, a very high value might be accredited to the services of a physician who can cure or alleviate disability. Similarly, a suit or legal criminal procedure against one might re ­ l4The Bowker Annual o f Library and Book Trade Information, 33d ed. (New York: R.R. Bowker, 1989); Statistical Abstract o f the United States: 1989, 109th ed. (Washington, D.C.: Bureau of the Census, 1989), 469. 15Rebecca Kellogg, “Faculty M embers and Aca­ demic Librarians: Distinctive Differences,” C&RL News 48 (November 1987): 602-606. sult in placing a high value on legal services. In contrast, lack of information may often be an in­ convenience, but seldom life threatening. T here­ fore, the services of a librarian may not be highly valued. The eventual effect this has on establishing the level of salaries offered to librarians results, to some extent, in the ultimate dismay librarians dis­ play about their pay. Concluding comments The concern over salaries so often expressed by librarians frequently lacks a frame of reference or a systematic analysis. W hen the concern is examined in term s of academic librarians com pared to fac­ ulty, and when the comparison is based on clearly stated economic principles, a b etter understanding of conditions emerges. Librarians’ salaries are relatively low. One might guess that librarians are paid salaries similar to those employed in disciplines that are restricted to the academic market. Unlike scientists, who can find competitive employment in government serv­ ice and industry almost at any point in a career, librarians are more restricted. That restriction is self-imposed, but produces gains that are probably b etter than the alternative. However, overall librar­ ian salaries may not be out of line with their faculty peers. F u rth er analysis using a data set containing information on faculty and librarians from many institutions would be needed to affirm this. How­ ever, the more im portant point is that librarians have some mechanisms open to them to do some­ thing about their perceptions, and it appears that they have. Librarians have taken two low-cost approaches to leveraging salaries upward. Given control of the credentials sought during searches, librarians have required minimum credentials. That requirem ent serves as a barrier, which restricts the supply of individuals available to hire. Given a m eans through their professional association to determ ine a standard on working conditions, librarians have leveraged salaries still further. It is an interesting mental exercise to com pare these techniques to those of other professions. Lawyers and account­ ants, typically, have invoked standard examinations to restrict entry into their professions. Beauticians restrict entry through licensing. O ther trades re ­ strict entry and bargain for wages through govern- m entally recognized agents. All o f these a p ­ proaches require substantial investments in legal mechanisms. Librarians appear to have been very creative in making use of low cost mechanisms to achieve similar results. W hether this is all that can be done to elevate wages remains to be seen, but it appears to be clear that the profession has not accepted salary conditions without making an ef­ fort to improve them. The So cia l It’s the one compact disc Sciences Citation to use for instant access to the bibliographic data from 1,400 international social sciences Index Compact journals — plus selected data from over 3,100 international Disc Edition. science journals. You’ll retrieve every article pertaining to your research topic, no matter what the discipline. Unprecedented Power The SSCI® CDE is the one compact disc with Related Records, the unique retrieval feature that links and displays bibliographic records that have references in common. So when you retrieve one record, you’re quickly led to additional records that are related to your search — items you’d never find any other way — with a single touch of a key. Only ISI makes this possible. See for yourself why the SSCI CDE stands apart from the crowd. To reserve a free trial copy, call toll-free: 800-336-4474, operator R318. Or write to one of the addresses below. Institute for Scientific Information® 3501 Market Street, Philadelphia, PA 19104 U.S.A. The SSCI CDE. Make it the one European Branch: 132 High Street, Uxbridge, Middlesex UB8 1DP, United Kingdom Telephone: +44-895-70016 S-12-6100 compact disc you use.