Home page for digital-law-online.info

- Table of Contents -

Introduction to the online version

Chapter 1 – The Commission and Its Recommendations

Chapter 2 – The Establishment, Mandate, and Activities of the Commission

Chapter 3 – Computers and Copyright

Chapter 4 – Machine Reproduction – Photocopying

- Recommendations of the Commission

- Recommendation for Amending One Area of the 1976 Copyright Act

- Recommendations Concerning the Five-Year Review of Photocopying Practices

- Recommendations to Publishers

- Recommendations to Government Agencies

- Provisions of the 1976 Copyright Act Affecting Photocopying

- CONTU Guidelines on Photocopying under Interlibrary Loan Arrangements

- Volume of Library Photocopying in 1976

- Means of Obtaining Permission to Make Photocopies or to Obtain Authorized Copies under the 1976 Copyright Act

- Interrelated Economics of Publishing and Libraries and the Impact of Copying Fees

- Legislation and Systems Relating to Photocopying in Other Countries

- Recommendations of Interested Organizations

- Effects of Future Technological Change

Chapter 5 – Summary

Appendix A – Summary of the Legislative History of Computer-Related Issues and the Photocopy Issue

Appendix B – Public Law 93-573 and Public Law 95-146

Appendix C – Commissioners

Appendix D – Staff

Appendix E – Lists of Witnesses

Appendix F – Alphabetical Listing of Persons Appearing before the Commission

Appendix G – Transcripts of Commission Meetings

Appendix H – Summaries of Commission-Sponsored Studies

Appendix I – Bibliography

Appendix J – Selected Provisions of the Copyright Act of 1976 and Copyright Office Regulations

Full table of contents

PDF version of the report

Picture of commissioners and staff

Final Report of the National Commission on New Technology Uses of Copyrighted Works

Chapter 4 – Machine Reproduction – Photocopying

Interrelated Economics of Publishing and Libraries and the Impact of Copying Fees

Many assertions were made concerning the economic impact of photocopying on journal publishing during the debates in the twenty-year effort of Congress to revise the copyright law, but little statistical or other factual evidence to substantiate these assertions was presented, either by librarians or by publishers and authors.

Librarians took the position that copying did not and would not significantly reduce the vol­ume of sales of printed publications, and that librarians strive to purchase little-used ma­terials rather than subject their patrons to the inconvenience and delay of interlibrary loan. In part, publisher and author concerns were not so {Page 66} much for the present as for the future, at which time they believed various means of photocopy­ing might become increasingly cheaper relative to the cost of acquiring printed publications. They also believed that they should secure some revenue from copying as a contribution to the “first-copy” costs of publications, such as edi­torial, typesetting, and business overhead.

General Relationship of Journal Publishing to Library Budgets

The most complete study on library and journal publishing economics and their relation­ship to one another is the 1975 Fry/White study, sponsored by the National Science Foun­dation, covering the years 1969, 1971, and 1973.243 Pertinent data were obtained through questionnaires filled out by a sample of U.S. scholarly and research journals and by a sample of academic, public, and special libraries.244 The periodicals included in the survey were ap­proximately 2,500 U.S. scholarly journals, of which about 150 furnished extensive usable data. The study provides data separately for four categories of journals by publisher: commercial publishers, professional societies, university presses, and other nonprofit publishers.

Library Acquisition Budgets – 1969-73

During the period 1969-73 when prices rose sharply for both periodicals and books, expendi­tures for periodicals and the number of sub­scriptions increased in all three types of libraries included in the Fry/White study. Although the total dollar expenditure for books also increased somewhat, the number of books purchased de­creased. These trends were more sharply accen­tuated in academic libraries-the largest pur­chasers of periodicals - than in public and spe­cial libraries. The median percentage increases in acquisition expenditures for periodicals and books from 1969 to 1973 in large libraries of all three types are displayed in Table 8.

Table 8 – Percent of Increase in Acquisition Expenditures, 1969-73

By 1973, large academic libraries, which had allotted 67 percent of their acquisition budgets to books and 33 percent to periodicals in 1969, were allotting 54 percent of the acquisition budgets to books and 46 percent to periodicals. The overall percentage increases in the number of periodical subscriptions by large libraries during the same 1969-73 period were: academic libraries, 18 percent; public libraries, 22 per­cent; and special libraries, 6 percent.

Borrowing and Photocopying through Interlibrary Loan

All types and sizes of libraries increased their borrowing of periodicals through interlibrary loan during the 1969-73 period. The median number of interlibrary loans and the percentage increases for large academic libraries, medium-sized public libraries, and large special libraries is delineated in Table 9.

{Page 67}

Table 9 – Median Number of Interlibrary Borrowings for Libraries

Effect of Photocopying on Periodical Subscriptions

Despite the frequent debate concerning whether library photocopying from journals sub­stitutes for subscriptions, little information has been available to resolve it. The responses that Fry and White received when they asked librar­ians whether photocopying had this effect on subscriptions are displayed in Table 10.

Table 10 – Effects of Photocopying on Periodical Subscriptions

(Calculated by Percentage Points)

While most libraries reported their purchases of journals were unaffected by photocopying practices, a substantial minority said otherwise. Although many libraries reported increasing subscriptions, the net effect on subscriptions seemed clearly negative in this study. Evidence from another study, however, indicates that in the state of Minnesota the total number of periodical subscriptions by fifty-three academic libraries in the MINITEX system increased by about 25 percent from 1970/71 to 1976/77 fol­lowing the establishment of this statewide inter-library loan network in 1971.245

Scholarly Journal Subscriptions and Library Budgets

The Fry/White study also analyzed the eco­nomics of U.S. scholarly journal publishing and the significance of the library market for such journals. Libraries constitute the largest market for scholarly journals. These journals are also the type of periodical most copied in or by libraries in relation to the size of circulations. The universe of 2,459 U.S. scholarly journals surveyed by Fry and White was broken down as shown in Table 11.

Table 11 – Breakdown of Number of U.S. Scholarly Journals by Type of Publisher

The journals responding to the questionnaire reported that the number of copies circulated increased in the period 1969-73, but generally this was the result of increases in the number of foreign and institutional subscribers and de­creases in the number of individual subscribers. Table 12 presents data by type of publisher.

The figures in Table 12 are consistent with the general trend in the Fry/White survey data which showed an increase in periodical subscription {Page 68} by libraries from 1969 to 1973.246 These journal subscription data do not Corre­spond exactly with the library data because, among other reasons, U.S. libraries subscribe to foreign journals and serial materials other than scholarly journals.

Table 12 – Percentage Changes in Circulation by Type of Publisher

The data also show that U.S. institutions-principally libraries-account for about one-third of the circulation of scholarly journals pub­lished by commercial presses, university presses, and other nonprofit periodicals, and for over 20 percent of the circulation of journals of pro­fessional societies which provide subscriptions to their own members as part of general mem­bership fees. The complete data on the propor­tion of subscriptions by type of subscriber are shown in Table 13.

Table 13 – Percent of Circulation Distribution by Type of Publisher, 1973

The publisher data displayed in Table 14 indicate that, in general, the number of journals that had differential (usually higher) subscrip­tion rates for institutions and libraries increased markedly from 1969 to 1973.

Table 14 – Percent of Journals Having Institutional and Library Subscription Rates, 1969-73

Table 15 shows that, in general, the institu­tional (usually library) subscription rates, when they exist, increased substantially more than the individual subscription rates in the period 1969-73.

Table 15 – Increases In Individual and Institutional Subscription Rates Median Subscription Rates

Taking all of the data from Table 15 into consideration, it is clear that libraries and other institutions provided an increased share of the revenues of scholarly journal publishers in 1973 as compared with 1969. Institutions were buy­ing a larger number of subscriptions as well as an increased proportion of all subscriptions. In many cases, they were also paying institutional subscription prices which increased more than the subscription price for individuals. Unfortu­nately, direct data on the proportion of total revenues derived by scholarly journal publishers from institution and library subscriptions are not provided in the Fry/White study. However, for many individual journals (except for society journals with high proportions of individual subscriptions) as well as for classes of journals, one-half or more of total subscription revenues must have come from U.S. libraries or other institutions. If foreign sales are added (and these are predominantly to institutions), this proportion is still higher.

{Page 69}

The Fry/White study also collected data on the profitability of U.S. scholarly journal pub­lishing. Statistically, this is the weakest part of the study, because only ninety-two journals pro­vided financial data. The data are summarized in Table 16.

Table 16 – Operating Income As A Percentage Of Total Revenue

Operating income is defined in Table 16 as “all revenue minus costs of sales and operating expenses. It does not take into account such items as interest paid or received, capitalized expenditures, or taxes.” Thus, for commercial publishers in 1973, the net profit after corporate income and other taxes would be about 6 per­cent of total revenue. The society, university press, and other nonprofit publishers would not, of course, be subject to federal and state corpo­rate income taxes. Although libraries in the 1969-73 period were subscribing to more schol­arly journals and paying higher prices for sub­scriptions, especially when they had to pay in­stitutional subscription rates, the net effect was not a windfall for the publishers of scholarly journals. Fry and White concluded that the price-budget imbalance did not result from ex­cessive returns to publishers. Commercial pub­lishers had profits no better than average, and societies had barely enough capital to launch new journals. They described the condition of university press journal publishing as disastrous.

Estimates of Possible Additional Costs to Libraries for Copying Fees

Using the estimates made from the King Research data in this report247 on the number of photocopied items from serials that would require authorization, and assuming a certain average level of photocopying fees, it is possible to arrive at estimates of additional annual costs to libraries resulting from copying fees under the new copyright act. These esti­mates are valid only if libraries do not change their copying practices as a result of the 1976 Act. It is likely, however, that libraries may alter their practices. The analysis uses the three cate­gories of copying in the King study (interlibrary loan, local use, and intrasystem use) and breaks down the copying by type of library. These amounts are then compared with the available data on total annual library expendi­tures for library materials to arrive at some esti­mate of the comparative magnitudes of these new costs as compared with the existing acquisi­tion costs for library materials. In the absence of comprehensive statistical data at this time, an average copying fee to publishers per article of $1.25 will be assumed. This is higher than the {Page 70} $.60 paid by the Institute for Scientific Informa­tion (ISI) and the $.50 paid by National Tech­nical Information Service (NTIS) under direct contracts with publishers. It may, however, be less than the weighted average price which will be paid to the Copyright Clearance Center. The 1977 Fry/White/Johnson study on journal publishing indicated that some 53 percent of the journals responding to a question relating to ap­propriate fees to be paid by agents or clearing­houses for copying articles set $1.00 or less as an appropriate fee.248

Copying for Interlibrary Loan

Table 3 and the discussion following pre­sented estimates of the volume of copying of periodicals for interlibrary loan for items not more than five years old that would not be exempt either under the CONTU guidelines, the exceptions in section 108 for replacement of copies, or section 107 for classroom use. Table 17 shows, at an average of $1.25, the additional annual costs to libraries.

Table 17 – Additional Annual Costs for Copying for Interlibrary Loan

($1.25 average copying fee)

The estimates in Table 17 may in some re­spects overstate the additional costs because they do not consider the following: (1) libraries reaching their limit of five copies for a title might subscribe to the journal,249 or tell patrons that their requests could not be met, or that they charge an additional fee for copying; (2) many journals may adopt more liberal copying policies than is required by law; and (3) authorized royalty-paid copies might cost the borrowing library the same as or less than conventional interlibrary loans.

Data compiled from a special survey of interlibrary loan charges by members of the Asso­ciation of Research Libraries in 1976 showed that the libraries in this group that charged for photocopies had a weighted average price of $3.50 for a ten-page article – excluding those cases in which special lower rates were charged to libraries in the same state (or consortium), or in which interlibrary loans were subsidized by the state. If this $3.50 figure is taken as a base and there is added to it an internal borrow­ing cost of $6.00, the total average cost to the borrowing library becomes $9.50. The $6.00 fee is selected as an internal borrowing cost; this is somewhat less than the average of the internal borrowing costs in 1977 in the three libraries for which such costs were calculated in the Palmour study contracted for by the Commis­sion.250

{Page 71}

These combined costs compare with an out-of-pocket direct cost to a borrowing library of securing a ten-page article from ISI of $3.50 (higher with special services), the $4.00 - $6.00 cost of securing a single copy from University Microfilms, and the price of a copy through NTIS, which will vary somewhat, but may aver­age $7.00. These comparisons do not take into account the internal costs to the lending library, over and above the fees charged.

Copying for Local Use

Table 4 presented estimates of the number of copies for local use by type of library re­quiring authorization. Applying the same as­sumed $1.25 average copying fee, Table 18 shows the additional annual costs for various types of libraries.

Table 18 – Additional Annual Costs for Copying for Local Use

($1.25 average copying fee)

As in the case of copying for interlibrary loan, the figures in Table 18 may in some re­spects be an overstatement of additional costs for some (but not all) of the reasons mentioned for interlibrary loan copying: (1) patrons might be informed that copies could not be made, or that they would have to pay an additional copy­ing fee; (2) many journals may adopt more liberal copying policies than are required by law; and (3) many older issues of journals will be out of copyright because they were not re­newed for a second term. On the other hand, the estimate for academic libraries may be low, because the King data on classroom use included all copying for classroom use, not copying for classroom use permissible under the educational copying guidelines.

Copying for Intrasystem Use

Table 5 presented estimates by types of library of the number of copies made for intrasystem use that would require authorization. Applying the same average $1.25 copying fee, Table 19 shows the additional annual costs for various types of libraries. These figures may be an over­statement of additional costs for the same rea­sons given in the discussion of copying for local use.

Table 19 – Additional Annual Costs for Copying for Intrasystem Use

($1.25 average copying fee)

Estimates of Total Additional Costs for Libraries

Based upon the above discussion and calcu­lations, the estimated costs for various types of libraries may be aggregated as shown in Table 20. Data from the National Center for Educa­tion Statistics (NCES) on the total expenditures of three types of libraries for library materials are also included.

Table 20 – Possible Annual Additional Costs to Libraries in Copying Fees for Periodicals as Compared with Expenditures for Library Materials

There are no reliable data on expenditures for materials by special libraries. Recently, NCES has contracted with the Special Libraries Asso­ciation for a preliminary study of special libraries in commerce and industry, including the ex­penditures for materials. The results of this study may be available before the end of 1978.

{Page 72}

Table 20 has some unexpected aspects. The net estimated increased costs for academic li­braries constitute an insignificant percentage of the current expenditures of these libraries for materials. The estimated copying fees of almost 3 percent of public library material expenditures constitute a very much higher percentage of ma­terials expenditures than for academic libraries, but the dollar amounts are not large.

The special library estimates may reflect the amount of multiple copying done in many of these libraries for their research, professional, and executive personnel. The federal agency library estimates are not particularly surprising, given the nature of most of their operations, which are more similar to special libraries than to either of the other two types of libraries.

If these estimates of copying fees are approxi­mately correct, the impact of photocopying on academic libraries as a class would be minimal and should not present any significant budget problems. Copying fees, which would go mostly to journal publishers, would not be great enough to do more than accentuate very slightly the trend in academic libraries of spending more of their acquisition funds on serials and less on books. For public libraries, the effect of photo­copying would be proportionately much greater, with copying fees amounting to almost 3 per­cent of total acquisition budgets in 1974. Since the bulk of the copying fees would be paid to periodical publishers, in the case of public libraries this might modestly accentuate existing trends of shifting funds from book to periodical purchases.

For special libraries, at least two different situations exist: one for libraries in business and industrial establishments or such related orga­nizations as trade associations, and the other for special libraries in nonprofit organizations. In the first category, information is used to increase the revenues or to reduce the costs of the busi­ness. Copying fees would also be a tax-deducti­ble cost of doing business. As compared to all other costs of doing business, copying fees would be small. Special libraries in nonprofit organizations are extremely varied, and it is difficult to state generally the impact of copying fees on their operations, especially since no statistical data exist on either their costs of op­erations as a class or their expenditures for periodicals, books, and other materials.

The federal library situation is somewhat similar to that of libraries in business and in­dustrial establishments. Information is used in carrying out the work of federal agencies. Copy­ing fees would represent another operating cost and should not, in general, be of such a magni­tude compared to other operating costs as to present unmanageable budget problems.

Potential Copying Fees Compared with Publishers’ Revenues - Periodicals

Data on the total revenues from periodical publishing are collected approximately every five years by the Bureau of the Census in the Census of Manufactures and are estimated each year based on a sample survey in the Annual Survey of Manufactures. Unfortunately, the Bureau of the Census divides periodicals into only four classes: farm periodicals, specialized profes­sional and business periodicals, general periodi­cals, and other periodicals (excluding shopping {Page 73} news, directories, and catalogs). Specialized pro­fessional and business periodicals, with esti­mated total 1976 revenues of $407 million from subscriptions and $946 million from advertise­ments, are those most likely to be photo­copied.251 It is clear that this broad category of periodicals is very different from the 2,459scholarly periodicals surveyed in the 1975 Fry/White study, the 1973 total annual revenues which were estimated at $170 million, of which less than 10 percent was from advertising. If we compare total annual copying fees for pe­riodicals, estimated to be about $10 million, with the total 1973 revenues of the scholarly journals surveyed by Fry and White, revenues from copying fees appear to be a minor but still significant source of revenue for some of these journals. While not comparable with revenues from subscription charges, income from author­ized photocopying could be, in some cases, more significant than such current sources of revenue as advertising, page charges, or a variety of sub­sidies.252

The Economics of Book Publishing, the Library Market for Books, and Photocopying

Libraries account for a very much smaller proportion of the total sales of U.S. books than they do of sales of scientific, technical, and scholarly journals, and the kinds of books which are photocopied in libraries fall into a few limited classes. The most copied classes of books and the industry estimates of total dollar receipts in 1977 of U.S. publishers (including exports) for these classes were as follows:253

   Technical and scientific                          $266.8 million

   Business and other professional              195.2 million

   Medical                                                97.0 million

   University press                                                         53.5 million

The library and institutional market is par­ticularly important to university press books, constituting well over one-half of university press sales within the United States.254 The King study clearly indicates that in 1976 the volume of photocopying from copyrighted books in libraries was considerably less than the volume of copying from copyrighted serials. The respec­tive proportions were: 70 percent of the items copied in libraries were from serials, 24 percent from books, and 6 percent from other copy­righted materials. These data are consistent with data from other studies and also are consistent with the general practice of libraries, which is to lend the physical book to their patrons for local use and also for interlibrary loans and intrasystem loans rather than to make photo­copies. The bulk of the copying from books in libraries has probably occurred in two ways: (1) the library patron may make a copy of some part of a book on a coin-operated ma­chine; and (2) library employees or others in educational institutions may make copies to place chapters or other portions of books on reserve for the use of students, or the instructor may have copies of portions of books made for class­room use. Because of the ambiguity of the 1909 Act, there also had been a certain amount of copying to create anthologies or substitute text­books by putting together photocopies of chap­ters of books and periodical articles for use in specific courses.

Congress dealt with copying for educational use by including in the House Report the edu­cational copying guidelines.255 These guidelines place definite limitations on the amount and character of copying for teachers and for class­room use that may be regarded as fair use under section 107. It is unlikely that the educational copying permissible within these guidelines will {Page 74} have much adverse effect on the economic via­bility of book publishing.

The amount of copying from books that is permissible under the 1976 Act, either as fair use under section 107 and the educational copy­ing guidelines or under the exemptions in sec­tion 108 for library copying, would seem to have no appreciable effect on the economics of book publishing at this time or in the next few years. There will undoubtedly be some copying of books in violation of the 1976 Act that will substitute for the purchase of books. The amount of such copying is probably small, however, and is inhibited by cost factors, in particular, the current frequently higher per-page cost of making copies as compared with the cost of buy­ing the book. Furthermore, there are whole cate­gories of books for which photocopies are not acceptable substitutes for the original product, such as paperback, book club, and art books.

The book publishing industry, although not highly profitable as compared with some other industries, has been fairly stable during the past few years. There has been a growing dollar volume of sales, much of which is a reflection of inflation and higher prices, with little or no increase in the number of copies of books sold. There is no question, however, that a problem exists with respect to the production and sale of scholarly books, most of which are now pub­lished in the United States by university presses. This problem is primarily due to inflation, lim­itations on the amount of su