IFLANET home - International Federation of Library Associations and InstitutionsAnnual ConferenceSearchContacts

61st IFLA General Conference - Conference Proceedings - August 20-25, 1995

Access versus Ownership: How Real an Alternative Is It?

Maurice B. Line, Library Consultant, Harrogate, UK


ABSTRACT

The assumption that access is to be preferred to ownership as a matter of policy needs to be questioned. Browsing and serendipity are lost in the access model. Access is better for periodicals tha n for monographs on the criteria of speed of supply, reliability and ease of use, but for both it is generally inferior to on-the-spot access. If cost were the only criterion, the current relative c osts of access and ownership, which favour access, may change substantially as and when access becomes electronic. Alternative forms of control and publication of research material have advantages a nd disadvantages. Ownership has limits: it can never approach comprehensiveness. Ownership of and exposure to a wide range of current material should be combined with access to older material. A s trong case could be made for larger acquisition funds in view of the coming emphasis on self-directed learning.

PAPER

Introduction

It has almost become established wisdom that the future for libraries lies in a policy of access rather than ownership. Librarians have either ceded the case for extensive acquisitions, or positivel y argued against it (e.g. Ferguson & Kehoe, 1993; Widdicombe, 1993), though there are exceptions (e.g. Truesdell, 1994). There are several arguments adduced to support this belief in access: that su ch a policy is unavoidable because much if not most material will be made available in the future in electronic form; that until this happens access will become easier and easier as more and more mat erials are produced in electronic as well as printed form or are digitized for purposes of transmission; that even now it does not matter much if an item is not in the stock of the library, because i t can easily be obtained from elsewhere; and that access is to be preferred on economic grounds, which is a vital consideration because funding will continue to be tight.

The Need for Exposure

While increased reliance on remote access may be necessary when funds are insufficient to maintain previous levels of acquisitions, it is another matter to accept it as a principle. One argument aga inst doing so is that the ability to browse is greatly impaired in the access model. Numerous studies have shown that even in the hard sciences users find much valuable material by browsing and sere ndipity. These two activities should incidentally be distinguished: browsing involves searching for material of relevance to one's subject of interest, often without precise subject terms under whic h to search, while serendipity is finding interesting material by accident. (Browsing has been compared with looking for a needle in haystack, while serendipity is looking for a needle in a haystack and finding the farmer's daughter). Both require exposure to the actual material. In the humanities the case for exposure is stronger still, even when the user knows precisely what he/she wants, s ince catalogue entries reveal little of the contents of monographs, and it may be necessary to look through many monographs of potential interest before finding what is wanted. Admittedly, some libr aries deny users this facility by putting much of their stock in closed storage, but in open access systems there is no doubt that browsing is almost universally practised and is much appreciated by users. Items that are recognized as interesting or relevant only when actually seen help researchers to extend their range of knowledge, make links between disciplines, and gain new ideas. Exposure to current material is especially important, since this enables researchers and teachers to keep abreast of new developments on and beyond the fringes of their own disciplines. It is after all cont inued exposure to such a range over many years that has given existing researchers the concepts and words whereby they can access databases at all.

It may be argued that text can be browsed on computer screens. Electronic access to the text of periodicals already offers online browsing. However, experiments to date have not held out much prosp ect that scanning screens can be an adequate substitute for scanning the printed page - the screen page is still too small, the process is much slower, scanning for prolonged periods is hard on the eyes, and even laptop computers are not as portable as printed materials (also, battery life is not very long, so that you have to be near a socket if you want to use them for long at a time). It is difficult to see how there can be any good alternative to exposure to a wide range of printed material.

There are now several CAS-IAS (Contents Access System - Individual Article Supply) systems, which provide access to the contents pages of numerous periodicals, and offer also supply of the articles l isted; examples are OCLC's ContentsFirst, CARL UnCover and the British Library Document Supply Centre's Inside Information and Inside Conferences (Leach & Tribble, 1993; Brown, 1993); but they cover few lower use periodicals, and article titles are only a partial guide to contents.

For monographs, the day of routine online access to text seems to be a very long way off, and in any case greatly improved bibliographic control would be needed to mitigate the loss of direct browsin g. Tables of contents of monographs are becoming available, but these often do not reveal very much. Much better would be extensive summaries; it is difficult to see this sort of provision being ma de generally, though the records produced by Book Data Ltd. for many English-language monographs go some of the way. Even this is no substitute for being able to look at the monograph itself. How m any of us buy books for our personal use in bookshops after merely looking at the title and contents pages? In fact, those who advocate access in preference to ownership admit when pressed that they are thinking only of periodicals, and then mainly of scientific and technical periodicals.

Relative Effectiveness of Ownership and Access

Should the decision to own or access be made on the basis of relative cost-effectiveness of access, or largely on effectiveness alone? The main criteria for effectiveness are speed of supply, reliability (probability of getting an item from the source or sources approached), and ease of use (Line, 1988). Costs are of course important, but varyin g degrees of prominence may be attached to them; I will return to them later. At present, access to "copiable" items like periodical articles satisfies all the above criteria much better than access to "non-copiable" items such as monographs. It is also better for scientific periodicals than for humanities periodicals. The reason is not that it is any longer much easier to locate periodicals than monographs - online access to union lists is changing that - but that the chance that a wanted book will be available immediately is lower. This too is improving as a result of systems whereby requests can be switched rapidly from libraries that cannot meet requests to ot her libraries, but this does not overcome the common problem that the monographs most likely to be requested are those that tend to be most heavily used within the libraries to which requests are sen t. This problem scarcely arises with periodicals, since they are not lent out by many libraries, and making a photocopy of an article does not deprive local users of the issue in question. Access t o monographs therefore suffers as to both reliability and speed of supply. Another reason why speed of supply is poorer for monographs is that photocopies can be transmitted by fax, and even if they are sent by mail they go by letter post, which is generally faster than parcel or packet post. It is most unwise for a library to offer access rather than ownership to any user who needs a monograp h at all quickly.

Access then is far more satisfactory for periodicals than for monographs. The differences will be increased rather than diminished by electronic access. In view of the differences, it is extraordin ary that most academic librarians keep periodical subscriptions going as long as they can at the expense of monograph purchases. Either they do not believe what they say about access as a real alter native, or they are obliged (by faculty) to act irrationally; there is a third possibility, which is that they are not aware of what they are doing.

Whatever the relative merits of access to periodicals and monographs, remote access to both is inferior to on-the-spot access. This is not to say that local availability does not leave a great deal to be desired. Availability studies in the UK and USA have shown not only that only about 70% of material wanted by researchers is held by their library, but that overall only 60% of the items actua lly held by academic libraries are available when they are wanted (Mansbridge, 1986; Chaudhry & Ashoor, 1994). The main reason for unavailability of held items is that they are in use, so that the 4 0% that are not available when wanted tend to be the most popular items, not those needed for research. Interlibrary access is of little use for such items, since they are mostly books needed quickl y by students. The only solution is to acquire many more copies, but, since the total library budget is limited, this leads to a reduction in the number of separate items acquired and so deprives re searchers.

Relative Costs of Ownership and Access

Relative costs alone are an inadequate criterion for deciding whether a local library should buy or borrow because of the importance of exposure and browsing, and because potentially useful materials that are not seen are often not identified and used. I have also argued that ownership scores higher than access on the main criteria. However, costs cannot of course be ignored.

One problem with comparing the costs of access and ownership is that the costs of access are not stable. Indeed, there is no standard cost for national access (for the sake of simplicity I am ignorin g international access, where the uncertainties are much greater). In more and more countries it is becoming accepted that the marginal (direct) costs of interlibrary supply should be paid for by th e requesting library; in very few libraries are supply and demand in balance, and the net suppliers are no longer prepared to carry the cost of satisfying an ever increasing volume of demand. In som e countries supply is partly subsidized (e.g. by excluding the cost of staff), but this is usually a temporary stage on the way to full direct cost recovery. Most libraries make different charges fo r loans and photocopies, partly because they may be obliged to charge a minimum rate for photocopies to satisfy legal requirements. If the price charged for loans is lower, this does not of course m ean that the cost to the requesting library is lower, since loans, unlike photocopies, have to be returned to the supplier. Suppliers also often charge photocopies according to the number of pages ( a practice that is not very rational, incidentally, since the cost of photocopying a page is very small compared with the cost of retrieval, postage, etc.; the difference between supplying a photocop y of two pages and one of twenty pages is a good deal less than one tenth of the total cost). Prices for periodical articles also vary according to whether mail or fax is used; the latter may be twi ce as high. Commercial suppliers charge full cost recovery plus a (generally small) profit [1].

Whatever the situation today, we can confidently predict that charging of full direct costs will eventually become the norm for national interlibrary supply. While this uncertainty remains, it is v ery difficult to make any sensible comparison of costs; certainly such a comparison is impossible for loans at present. Where commercial suppliers are used for periodical articles, the cost could be anything from twice that of library suppliers to the same; if libraries charged full costs, these could actually be higher than those of commercial suppliers. If articles are transmitted from elect ronic versions by publishers or hosts, the cost may be higher, for the reasons given above.

There have been several attempts to calculate cost comparisons for local libraries in the case of periodicals. Estimates of the number of times a volume of a periodical has to be used before it is m ore economic to buy it range from five to ten. These estimates are of very little value, since the number obviously depends on the price of the periodical. Average prices in the UK range from £88 for a humanities periodical to £403 for a science periodical (1994 prices) (see Appendix 1). These averages conceal enormous differences in prices of individual titl es. In the case of monographs, the average price in the UK (January-June 1994) of a British academic book was £35.42 (ranging from £86.65 in botany to £14.55 in literary texts); the average price of a US book was $44.65 (ranging from $111.34 in surgery to $24.11 in sports and recreation) (Appendix 2). To these costs must be added those of selection, acquisition, processing (including and especially cataloguing), and shelf space.

There is yet another problem in making a cost comparison between acquisition and obtaining a loan or photocopy: the expected use has to be estimated. As I shall point out later, librarians are inacc urate when making positive selection decisions when acquiring monographs; they are likely to be less accurate still when deciding about uncertain cases. For periodicals it is rather easier, since use studies can provide data for those already acquired, while interlibrary access figures can provide data on those t hat are not. However, reliable use studies are not easy to carry out, and interlibrary access figures cannot show what use would be made of the periodicals if they were present in the library.

The best that can be done is to give one or two examples. Let us say that a monograph costs £35 to buy, and an additional £15 to process; if storage costs are ignored, the total cost of acquiring it is therefore £50. Let us also assume that charges are made for all loans, that an interlibrary loan costs in total £10. It then becomes more economic for a library to acquire if the expected use i s more than five. If no charge is made, the costs will consist of return mail and internal costs - perhaps £5.

As a second example, let us take a science periodical costing £300, to which may be added £40 in processing and binding (it has 12 issues a year and occupies two bound volumes): £340 in all. Let us also assume that the total cost of obtaining a photocopy of an article is £10. It is economic for a library to acquire the periodical if use per volume is 35 or more. It is not surprising that m any authors (e.g. Gossen & Irving, 1995) have concluded that it is not cost-effective (or for that matter necessary for other reasons) to continue to acquire low use periodicals.

The periodical example illustrates the very dubious utility of making cost comparisons of this kind. If they were applied strictly and acted upon, there would be few science periodicals left in libr aries (and in consequence there would be fewer published). The academic community would rightly object strongly to this, for its work would be very seriously affected.

There are in any case too many uncertainties for decisions to be made on the basis of costs alone. The vast differences in the price of individual periodicals and monographs, and the variable and ch anging situation with regard to charges for supply, mean that any calculation would not only be extremely complex but would remain valid for only a short time.

The issue of relative costs can be bypassed, and the problem solved from the library's angle, by passing on costs of access to users. If all direct access costs are charged to users, it makes econom ic sense to buy as little as possible and use access as much as possible. If this is carried to its logical conclusion, the library would buy nothing and access everything, though even the most arde nt advocates of access do not suggest this. For a library to charge users anything for access, other than perhaps a small fee to deter frivolous use, is to penalize them for poor selection or inade quate acquisition; it is no more logical than charging them for use of stock in the library. Such a "solution" ultimately defeats the purpose of the library.

The Future Economics of Access

While it may be more economic in strictly limited terms to access remotely rather than acquire a monograph or periodical that is wanted less than a certain number of times, this is true only a s things stand at present. When many libraries cancel a periodical, the price goes up, thus reducing purchases still more, and favouring access. But after a certain point the periodical ceases to b e viable as a publication, and so is not available at all in conventional form. Likewise, if the estimated sales of books of a certain type (say, archaeological monographs) fall below a certain leve l the books are not published.

Some argue that electronic access will overcome these problems, at least for periodicals. The possibility of 'books on demand' has also been mooted: monographs with an estimated small market could b e stored electronically and printed out speedily on demand. However, publishers (and hosts) have to recover their costs, and access costs for low use periodicals will not be low, even though the cos ts of publishing would be much reduced in an all-electronic system because the costs of paper, printing and binding would not be incurred. Publishers will not readily surrender the profits they are making from printed journals and monographs, and although they will welcome an additional source of income from the sale of individual articles (and are already beginning to enjoy such an income), th ey are unlikely to give up publishing printed versions unless they can make a greater profit from supply on demand. There is always the possibility that a publisher will offer items on demand at a l ower than cost price for a time to capture the market, but only the richest publishers would be able to do this for more than a few months. Elsevier Science Publishers have stated that electronic periodical subscriptions will not be cheaper than those of paper-based products, at least in the short term (Steele, 1995). One of the reasons why publishers are interested in electronic access is that it gives them much greater control over use; they can in fact set whatever prices they like within the limits set by the market. It could be that libraries are falling into a trap, though I doubt if it is one deliberately laid by publishers: they cancel periodicals until electronic access is the only means of access possible, and they are then entirely in the publishers' hands. The costs of a "books on demand" system are unknown, but they could not be low.

It may be argued that market considerations will make this scenario unlikely; for if prices of access are too high, the articles will not be used. There is undoubtedly some truth in this, but it has been shown time and time again that if an item is wanted badly enough it will be paid for; and if the library cannot pay for it the individual or the department will (the same departments that fight for their own budget and do little to plead for decent library funding). This is already happening; users are beginning to bypass the library if it cannot or will not supply what is wanted. An alt ernative market has been created.

Alternative Systems of Access and Supply

One proposal that has received a good deal of support in the USA and UK is that the storage and supply of articles should be taken out of (or never put into) the hands of publishers and become the re sponsibility of the institutions that fund the work on which the articles are based. Presumably there would be a sort of consortium of universities - or rather national consortia, since it is hard t o imagine such a system on any scale larger than a national one. The control and management of such a system (or systems) would be an awesome pro-spect, quite apart from the question of quality cont rol, which the present system provides through the refereed packages of articles we call periodicals. Indeed, bearing in mind the strong pressures in many countries to contract out to the private se ctor anything that can be contracted out, if universities did set up such a system it would seem to be a prime candidate for privatization.

One form of ownership, albeit under conditions - an intermediate possibility between electronic access to remote sources and conventional acquisition of paper periodicals - is the acquisition or leas e of CD-ROMs of full text. One obvious example is ADONIS, which makes over 600 mainly biomedical per-iodicals available on CD-ROM. This is a good deal cheaper than purchase of the paper versions, b ut its cost-effectiveness depends on how many of these titles are actually wanted; also, exposure and browsing are greatly reduced. A much better system would be one that allowed the acquisition of tailor-made CD-ROMs, containing titles selected to suit each library's requirements, but this would be far more costly.

Another possibility is to build up a collection of articles by keeping a digitized copy of every article acquired from elsewhere for a user. Quite apart from problems of copyright, however, the prob ability that the same article will be wanted more than once is remarkably low. This is not only true locally, as studies at Loughborough a few years ago showed; surprisingly, it appears to be true n ationally as well, and for this reason there is doubt about the virtue of scanning articles and maintaining a file for future use.

There are various possible forms of cheaper publication, all of which would make ownership easier. They include synopsis journals (which were tried some twenty years ago and failed because authors d id not like them), miniprint (reduction to a size that can still be read with the naked eye), and tabloid publication (periodicals on cheap paper of tabloid newspaper size, which might attract person al subscriptions; they would be published also in electronic versions for permanent storage). It is impossible to discuss all of these here; I merely plead for them to be looked at seriously (Line, 1992).

The Limits of Ownership

It has to be accepted that, however generous local funding is, it cannot ever be enough to meet the needs of users. It was shown earlier that local availability is often poor both on terms of the pr oportion of wanted materials that are held and access to those that are held. However, to improve the performance of the collection (as opposed to the availability of held items) by 10% would probab ly require an increase of some 30% or more in acquisitions: in other words, increases in acquisitions yield diminishing returns.

It has also been shown that however carefully libraries carry out selection, a high proportion of the books and periodicals acquired are rarely or never used. This is true even for books acquired fo r students; studies in three not very well funded academic libraries in the UK indicated that some 30% of such books were not borrowed in the years immediately after acquisition (in-house use was not studied) (Line, 1986). It must apply even more strongly to material acquired for researchers; the well known Pittsburgh study (Kent & others, 1979) is not the only one to show that much of the stoc k of academic libraries is never used. (The problem of inaccurate selection applies with much less force to periodicals than to monographs, since future demand for periodicals can be estimated with fair accuracy on the basis of past use.)

Monograph selection is partly an act of faith; and a lot of money is wasted by the purchase, processing and storage of unused books. The proponents of access could use this as an argument for their cause: since it is impossible to select accurately, a better service would be rendered to users if items were obtained as they are wanted and the 'wasted' money used for material that was actually wa nted. One could draw an opposite conclusion: that since selection is imperfect, and since much material is picked up by browsing, it is all the more important that there is a wide range of material from which users can select - an argument in favour of good acquisitions budgets. In any case, as Gorman (1994) points out in stating a case against the virtual library, the use per item in very lar ge collections is often very high; he estimates that annual usage of the University of California Libraries amounts to over 50% of the size of the collections (though this does not of course mean tha t 50% of the col-lections is used).

Access to Current and Older Material

There is a critical distinction, which has received too little attention, between access to and use of current material and older material. This distinction is much more pronounced in the sciences t han in the humanities, with the social sciences falling somewhere between. While it is certainly possible to get ideas and stimulus from older scientific literature, it is rare, and happens mostly b y serendipity; most of the literature has been superseded (though this is less true of the descriptive life sciences such as botany), and its bulk is so vast that browsing is virtually impossible. W hat is ideally needed is a combination of exposure to a great deal of current literature and speedy and efficient access to older literature. What most libraries offer at present is a selection of b oth older and current literature, though they nearly all also offer access through databases and document supply systems. The selection is becoming smaller as acquisition budgets fall further and fu rther behind inflation and literature growth. A sensible policy would be for libraries to discard most of their stocks of material more than, say, three or four years old, thus freeing large areas o f space, and provide on the spot far more current material. For this to be possible requires the existence of a publicly accessible comprehensive collection of older material, in an institution gear ed to rapid and efficient supply; such a collection need not be in the country of the enquirer, though it is faster and cheaper to obtain monographs (as opposed to periodicals) on loan from within th e country. The British Library Document Supply Centre, whose services are worldwide, is the nearest thing there is to such a collection (it has of course to acquire material currently, both because it is in heavy demand and in order to ensure that it is available when it is no longer current).

This last point - the need to keep material available when it is no longer current - is important. There is a danger that once published material is no longer in public hands it will cease to be av ailable some day, whether because the publisher goes out of business (as publishers are apt to do), or because he no longer wants to keep it available. There is no real reason why he should not, sin ce electronic storage is cheap enough, but he has no obligation to do so. There have been discussions on archiving material that is held only in electronic form, but so far no solution has been agre ed.

How would the economics of a system of current ownership and access to older material work? The very substantial savings in space would be capital savings, and while they would doubtless be much app reciated by parent bodies they would not solve the problem of current expenditure. Money could also be saved by avoiding the binding of most periodicals; since they would be disposed of after a few years it would be pointless to bind them. Stock main-tenance costs would also be reduced. These savings would free some money, but not a vast amount. So more money would undoubtedly be needed, for the purchase of more material and also (a much smaller sum) for the staff required to process it. But what is the alternative?

There are only two alternatives. One is for users to rely much more on remote supply, whether online or by the supply of photocopies or loans. As explained above, this is not only less satisfactory ; it may also be more expensive, no less so because the costs may not be borne by the library but by the individual or a university department. Because the costs would be visible, demand would proba bly be reduced; some of this reduction might be in "unnecessary" demand, but it is not easy to decide what is unnecessary in advance. The other alternative is worse still: that users are not served properly at all, and that their research or study is thereby seriously restricted if not damaged.

In fact, there would probably be a combination of both alternatives. More dependence on remote supply would prove more costly if users asked for as much as they would use if it were available on the spot; but they would not, simply because it was not available, and either they would not be aware of it at all or they would not make the effort to ask for it. It could be argued that if the y were not prepared to make this effort they could not be seriously interested; but experience does not suggest that reluctance to make an effort is always a good indicator of lack of need. Usage wo uld therefore be reduced, irrespective of any restrictions imposed on expenditure on remote access; and research and teaching would be affected as a result. If restrictions were imposed, the effect s would of course be greater.

Conclusion

Libraries will of course need both ownership and access. No-one wants to go back to the days when to request an item from another library was an ad-mission of failure. The questions to be answered concern the balance between the two, and the general policy to be pursued.

Access can serve as a reasonable substitute for holdings of older material, but not for much current material. Librarians have been too easily seduced by the rhetoric of the technological prophets, or have too passively succumbed to the pressures on them. Prophets are not always right (those that are tend not to be believed), and in this case I believe they are both wrong in their costings, an d either innocent or misleading in their unawareness of the needs of users and the likely effects if their visions became reality. Access will not necessarily be cheaper in the future, and it would be unwise to plan on the assumption that it will; and I believe it cannot substitute for some important functions of the printed page.

The whole rationale of universities in particular is to promote study and research; if they cannot do this effectively they are failing. It is not as if huge sums of money were at stake; in the UK, if expenditure on libraries went up from its present figure of less than 3% of university expenditure to 5%, it would transform libraries without serious damage to teaching or other areas of expendit ure (70% of university expenditure goes on academic staff). The cost of a well stocked library may be going up more steeply than inflation, but it is not in fact all that high. A country that claim s it cannot afford such libraries should be ashamed to call itself developed.

In fact, librarians could make a very powerful case for larger budgets. I have no doubt that the future of higher education (and possibly of secondary education as well) will be one of self-directed learning; economics will make it necessary, technology is making it possible, and the value of self-direction makes it desirable. Instead of libraries supporting teaching, we shall have teachers su pporting learning resources - which of course include libraries (Line, 1995). This is already beginning to happen. Another major trend is lifelong learning, which also favours expenditure on learni ng resources. Why are librarians not arguing this case?

I have in the past argued for "leaner, fitter libraries". There is no contradiction between this view, which I still hold, and the approach in this paper. Leanness does not mean emaciation, and fit ness requires the ability to do what is needed, which in the case of libraries is to supply what users need in the most efficient manner. Libraries need to shed the unnecessary fat of unused older s tock, not lose the muscle that current material gives them. A few seem to be in danger of combining bulimia - accumulating as much old material as they can lay hands on - with anorexia - starvation of current proteins and vitamins. Admittedly, the starvation is not self-imposed, as anorexia is; but it is no more agreeable to users for that.

1. Several recent American articles compare the performance of library suppliers with commercial suppliers (Miller & Tegler, 1988; Kurosman & Durniak, 1994; Pedersen & Gregory, 1994). No clear advantage in the use of either one or the other emerges from these studies, which show the expected large differences in cost and performance of commercial sup-pliers.
2. These are rather lower than the average figures calculated for US libraries, which are $18.62 for the requester and $10.93 for the supplier (Roche, 1993).

REFERENCES

Brown, David J. (1993). Current awareness and document delivery: the changing market. Serials, 6(3), November 1993, pp.29-39.
Chaudhry, Abdus & Ashoor, Saleh (1994). Comprehensive materials availability studies in academic libraries. Journal of Academic Librarianship, 20(5/6), November 1994, 300-305.
Ferguson, Anthony & Kehoe, Kathleen (1993). Access vs. ownership: what is most cost-effective in the sciences. Journal of Library Administration, 19(2), 1993, pp.89-99.
Gorman, Michael (1994). The "unused collection" myth. California Librarian, 4(10), October 1994, p.3.
Gossen, Eleanor & Irving, Suzanne (1994). Ownership versus access and low-use periodical titles. Library Resources and Technical Services, 39(1), January 1995, pp.43-52.
Kent, Allen and others (1979). Use of library materials: the Pittsburgh study. New York [&] Basel: Marcel Dekker, 1979.
Kurosman, Kathleen & Durniak, Barbara Ammerman (1994). Document delivery: a com-parison of commercial document suppliers and interlibrary loan suppliers. College and Research Libraries, 55(2 ), March 1994, pp.129-139.
Leach, Ronald G. & Tribble, Judith E. (1993). Electronic document delivery: new options for libraries. Journal of Academic Librarianship, 18(6), January 1993, pp.359-364.
Line, Maurice B. (1986). Can book selection be improved? [Review essay]. British Journal of Academic Library Management, 1(2), Summer 1986, pp.160-166.
Line, Maurice B. (1988). Measuring the performance of document supply systems. Interlending and Document Supply, 16(3), July 1988, pp.81-88.
Line, Maurice B. (1992). The publication and availability of scientific and technical papers: an analysis of requirements and the suitability of different means of meeting them. Journal of Docume ntation, 47(2), June 1992, pp.201-219.
Line, Maurice B. (1995). The road through chaos: the future role of the university library in the creation of knowledge. In: Collection and cooperation: essays on tradition and change presented to Thomas Tottie. Edited by Sven Hedberg. Uppsala, 1995. in press.
Mansbridge, John (1986). Availability studies in libraries. Library and Information Science Research, 8(4), October/December 1986, pp.299-314.
Miller, Connie & Tegler, Patricia (1988). An analysis of interlibrary loan and commercial document supply performance. Library Quarterly, 58(4), October 1988, pp.352-366.
Pedersen, Wayne & Gregory, David (1994). Interlibrary loan and commercial document supply: finding the right fit. Journal of Academic Librarianship, 20(5/6), November 1994, pp.263-272.
Roche, Marilyn M. (1993). ARG/RLG interlibrary loan cost study. Washington, DC: Association of Research Libraries & Research Libraries Group, 1993.
Steele, Colin (1995). Publish or perish / converge or die? The future of scientific inform-ation access (electronically). Information Management Report, February 1995, pp.1-4.
Truesdell, Cheryl B. (1994). Is access a viable alternative to ownership? A review of access performance. Journal of Academic Librarianship, 20(4), September 1994, pp.200-206.
Widdicombe, Richard P. (1993). Eliminating all journal subscriptions has freed our cus-tomers to seek information they really want and need: the result - more access, not less. Science and Techn ology Libraries, 14(1), Fall 1993, pp.3-13.

APPENDIX 1: Average periodical prices in UK (1994)

Humanities and social sciences                          £88.24
    highest: Psychology                                 £172.55
    lowest:  Sports & pastimes                          £34.75

Medicine £225.84 highest: Surgery, anatomy, physiology £327.92 lowest: Public health, nursing, general practice £100.89

Science and technology £402.55 highest: Biophysics, biochemistry, microbiology £785.24 lowest: Heating, lighting, ventilation £52.70

Source: Library Association Record, May 1994 ('based on a broad selection of periodicals made by Blackwell's Periodicals Division').

NB. Within the 'highest' categories, there are of course great variations between individual titles. ____________________________________________________________

APPENDIX 2: Average book prices in the UK (January-June 1994 figures)

UK books                            US books

overall average £35.42 overall average £44.65

highest: botany £86.65 surgery £111.34 chemical engineering £84.38 mechanical engineering £103.91 surgery £75.83 chemical engineering £103.64 lowest: literary texts £14.55 sports & recreation £24.11 nursing £16.22 literary texts £25.65 sports and recreation £20.65 general works £26.92

Source: Sumsion, John. Average prices of British academic books. January - June 1994. Loughborough: LISU, 1994. (LISU British Academic Book Prices Report, no.15)
Sumsion, John. Average prices of USA academic books. January - June 1994. Loughborough: LISU, 1994. (LISU USA Academic Book Prices Report, no.15)