Research Brief 3

Comprehensive Access to Off-Site Print Materials at Johns Hopkins University

In June 1996, within a program funded by The Andrew W. Mellon Foundation to encourage research on the economics of information, the Council on Library Resources (CLR) provided support to Johns Hopkins University for a project to analyze the economic feasibility of offering electronic access to materials in remote storage. Access was to be achieved through a combination of means-robotics, planetary scanners, software, and high-speed communications. This Research Brief summarizes the final report on the project by the principal investigators, G. Sayeed Choudhury, Head of the Digital Knowledge Center at the University’s Milton S. Eisenhower Library, and Prof. Ben F. Hobbs, Department of Geography and Environmental Engineering, the Whiting School of Engineering.


One of the most serious problems confronting research libraries is the lack of adequate space for their print materials. In November 1995, the Eisenhower Library at Johns Hopkins University opened an off-site shelving facility that will eventually provide high-density storage for almost 2,000,000 volumes. The facility currently houses more than 425,000 books, 250,000 microfiches, 31,000 microfilms, and nearly 3,000 boxes of archival/manuscript materials, and the total number of stored items is expected to grow by approximately 40,000 annually.

The stored materials are noted in the online catalog system at the Eisenhower Library, and requests for them are made in the Library on a form that is forwarded to the remote site. Items requested before noon can be delivered to the circulation desk later the same day; items requested after noon are usually delivered the following weekday (there is no service on the weekends). The average delivery time from the off-site facility to the campus-based user was estimated at 52.12 hours. As users have become more familiar with the request process and more knowledgeable about the materials in storage, the number of requests has risen steadily. There were 7,081 such requests during the remote facility’s first year of operation, at least 77% of them from Hopkins faculty members and students.

The stored materials are kept in boxes by size rather than by call number and are retrieved by staff using a forklift. The drawback of the facility, as of all such, is that it restricts direct access to the materials by users, who can no longer be browsers. The goal of the University’s Comprehensive Access to Print Materials (CAPM) project is to give users “real-time” electronic access to the stored materials through technological means. A fully operational CAPM system will make possible complete “remote control” of the materials: upon request, items will be retrieved from the shelves of the storage facility and delivered to a scanning area; there, a planetary scanner will produce digital images that allow users to browse the items and, if desired, print pages; finally, users may choose to return items to the shelves or request them for physical delivery.

Financial Analysis

The Library and the Whiting School conducted a successful technical feasibility analysis of CAPM. Then they undertook the financial analysis, which had two components: one sought to assign value to the potential benefits of implementing CAPM; the other sought to estimate the costs of various versions of CAPM, ranging from a hybrid system that would utilize both automated and human components, to a fully automated system.

The benefits that would accrue from implementing CAPM include : earlier access to materials (the requested documents would be available for viewing electronically within minutes); the convenience of accessing the content of materials through any networked computer (CAPM would end the current procedure of having to fill out request forms and return to the main library to claim items); and the ability to browse in requested items, print or download specific pages, and, in effect, enjoy real-time intellectual access, even from sites beyond the University.

Users would surely agree that these are valued benefits, but can a dollar value be put on the convenience? Library services are “goods” without clearly defined markets, and patrons quite reasonably have difficulty assigning dollar values to library services. As a consequence, contingent valuation methods, or other survey-based approaches, are often adopted to judge the value of these services. Economists assert that people’s time is valuable, and a reasonable proxy for the value of one’s time is one’s wage rate. Thus, if a procedure saves an individual an hour of time, one might argue that the value of the procedure to the individual is his/her hourly wage rate.

Using representative low, average, and high hourly wage rates, the Hopkins team calculated, in dollar terms, the total annual value of the benefits of earlier access at $208,900, $271,546, and $334,193. With 7,081 requests for materials in the year, the values per item were $29.50, $38.30, and $47.10, respectively. The results of other surveys suggest that these figures may be too high. Other factors suggest that they may be too low. This uncertainty caused the CAPM team to decide that it is perhaps inappropriate to assign dollar values to the benefits of implementing CAPM.

The Hopkins researchers then took a different approach to assessment. They adopted “a multiple-criteria decision-making graphical presentation” (MCDM) to analyze the merits of the CAPM system. MCDM methodologies define a set of objectives (which are either maximized or minimized) subject to constraints (budgets, for example). The framework assesses projects without relying on monetary metrics. The CAPM team contends that libraries, which do not have markets for their “goods,” are an ideal arena in which to apply the technique: the graphical presentation acknowledges the subjective nature of evaluating the benefits of library services and explicitly recognizes the unique contribution of the CAPM system. From their final report:

“Benefits and costs measured in monetary terms are essential for traditional cost-benefit analyses. However, a traditional cost-benefit analysis generally provides a mechanism for assessing the cost-effectiveness of alternate operating systems which accomplish the same task. While the CAPM might provide a more efficient delivery system for off-site materials, its greatest benefit lies in the ability to provide scholars, even outside of Johns Hopkins University, realtime intellectual access, independent of time and space, to off-site print materials-a feature which is unavailable with the current delivery system. If the CAPM system is implemented in other institutions with off-site facilities, this benefit can be extended to all scholars (with access to networked computers) for any off-site print materials. This point can not be overemphasized.”

The off-site Hopkins facility is an independent unit, and most of its costs are not shared with the Eisenhower Library. The total annual operating costs of the facility are estimated at $250,000, a figure that includes rent, property taxes, operations and maintenance, staff salaries, and transportation costs. Using this figure as the baseline cost, the investigators determined the additional costs for hybrid and automated versions of the CAPM system. They outlined two versions of CAPM for this analysis-one that relies upon human operators for item retrieval, placement on scanners, and page-turning, and a second that replaces the page-turning component with an automated system. They included no amortized research and development costs in the analysis because it was assumed that these costs would be met by external sources; they estimated only the additional operating costs.

For the first system, the feasibility team decided that three CAPM stations could be operated by two full-time staff members who would retrieve items, place them on the planetary scanners, and turn pages at the users’ request. With three shifts of operation each day, an additional four full-time staff members would be needed. Each planetary scanner and printer (for the CAPM station) costs approximately $25,000. Salary support for each of the four full-time staff members is approximately $35,000. Thus, a hybrid version of CAPM would require an additional $215,000 annually.

An automated version of CAPM would include a robotic page-turning device that would be purchased for an estimated $10,000 per station. In this case, it was proposed that one fewer staff member per shift would be required, but an additional three CAPM stations could be operated (for a total of six). Thus, the six planetary scanners (at $25,000 each) would require $150,000, and the two staff members (at $35,000 each) $70,000. Add the cost of six robotic page-turning devices (at $10,000 each), and the total cost becomes $280,000.

The Hopkins team proposed the following summary to show costs and benefits of CAPM in monetary terms (all dollars are 1997, and low, average, and high refer to assumed wage rates):

Table 1: Comparison of Costs and Benefits
System Annual Operating Costs ($) Additional Annual
Cost ($)
Additional Annual Benefits ($)
Current System 250,000 0 N/A
Hybrid CAPM 465,000 215,000 208,900 (low)
271,546 (ave.)
334,546 (high)
Automated CAPM 530,000 280,000 208,900 (low)
271,546 (ave.)
334,546 (high)

However, recognizing how uncertain the dollar values are, they had recourse to the MCDM representation. An MCDM presentation highlights the implicit trade-offs in any decision-making process and reflects the difficulty, and perhaps the impossibility, of translating these trade-offs into dollar values. The Hopkins investigators used value-path diagrams to illustrate the trade-offs involved in implementing CAPM.

The MCDM Representation

In the language of MCDM, the problem might be stated as follows: The ideal or “optimal” solution to the problem of accessing remotely stored materials is one that minimizes access time and costs and maximizes convenience, the ability to browse, and real-time intellectual access to off-site print materials.

Implementing CAPM yields the following benefits, of which the last is perhaps the most significant:

  1. rapid access to materials
  2. convenient access to contents from remote locations
  3. the ability to browse contents
  4. the capacity to provide users, even outside the university, real-time intellectual access, independent of time and space, to off-site print materials.

With each system there are the operating costs outlined earlier. (Note that the access time for CAPM was estimated at 12 minutes.)

In Table 2, the third column offers a metric for comparing the convenience of delivery: the current system is set at 1, and the author makes a subjective assessment, which others might question, that either of the CAPM systems would be four times as convenient.

Table 2: Benefits in Terms of Objectives
System Annual Operating Costs ($) Convenience Ability to Browse Real-Time Intellectual Access Number of CAPM Stations Access Time (hours)
Current 250,000 1 0 0 0 52.12
Hybrid CAPM 465,000 4 1 1 3 0.2
530,000 4 1 1 6 0.2

The ability to browse and “real-time intellectual access” are absent from the current system and are therefore assigned values of 0. The CAPM systems are assigned a 1 to reflect the presence of these capacities.

The annual operating cost of $250,000 and the access time of 52.12 hours are taken as baseline figures, and figures in the first and sixth columns of Table 2 are divided by them, as reflected in Table 3.

Table 3: Normalized Benefits in Terms of Objectives
System Annual Operating Costs ($) Convenience Ability to Browse Real-Time Intellectual Access Number of CAPM Stations Access Time (hours)
Current 1.00 1 0 0 0 1.00
Hybrid CAPM 1.86 4 1 1 3 0.00
Automated CAPM 2.12 4 1 1 6 0.00

The team plotted the values from Table 3 with parallel axes to create the value-path diagram shown as Figure 1. Recall that the objectives are the following: to minimize annual operating costs and access time, and to maximize convenience, the ability to browse, real-time intellectual access, and the number of CAPM stations.

Figure 1

The results of the study persuaded the University to move forward with the implementation of CAPM and to seek funding for a prototype system at the off-site storage facility.

Research Briefs are occasional papers published by the Council on Library and Information Resources (CLIR) to describe the outcome, or the current status, of projects undertaken within its programs. CLIR encourages duplication of these papers and requires no permission for their further distribution.