College and Research Libraries Economics of the Scholarly Journal David W. Lewis This paper considers the economic nature of the scholarly journal from a theoretical perspective and concludes that it is what economists call a natural monopoly. Natural monopolies exist when the average price of the good falls over the range of demand, and unless a subsidy is provided the good will not be produced in the quantity that provides the most social benefit. The natural monopoly model of the scholarly journal sheds light on the issue of dual pricing and explains how scholarly publishing can be a highly profitable enterprise. Because subsidies should be easier to implement in electronic systems, this alternative may provide a more effec- tive means of scholarly communication. ibrarians and the publishers of scholarly journals have a long history of disagreement over prices. Librarians feel ex- ploited, and publishers misunderstood. Neither side seems to be able to see the other's point of view .1 This lack of com- prehension occurs, at least in part, be- cause librarians are not knowledgeable about the economics of scholarly journal publishing. If librarians are to make rea- soned and reasonable policy decisions on the distribution of scholarly information, the situation must change. The scholarly journal is only one means of distributing scholarship, but if we understand its eco- nomics, we can apply our insights across the board. This article will explore the eco- nomics of the scholarly journal and along the way will consider dual pricing and the changes electronic systems may bring. We will begin by examining the schol- arly journal's production characteristics and the nature of the scholarly journal's demand as a product in the marketplace. To do so, we will use a simple example to explore these characteristics and the inter- play that results among libraries, scholars and publishers. As is the case whenever simple models are used to portray a com- plex reality, some of the detail and texture will be lost, but in compensation we will be given the opportunity to see clearly re- lationships that might otherwise elude us. SUPPLY Different goods are made in different ways. To describe the differences, econo- mists develop production functions that explain how, for the particular product, inputs are turned into outputs. The pro- duction function for scholarly journals is twofold. There is a large, up-front fixed cost-what publishers call the "first copy cost.'' This expenditure is necessary to se- lect and edit articles, to lay out pages, and to maintain subscription lists and a distri- bution system. It must be paid regardless of how many copies of the journal are sold. The cost of printing and distributing each unit after these setup costs have been paid is relatively small. Let us assume for our example that the fixed costs for a year ' of production of a hypothetical journal are $10,000 and that the per-unit costs are $10 per subscription. The total cost of produc- David. W. Lewis is Head, Research and Infonnation Services at the Homer Babbidge Library, University of Connectzcut, Storrs, Connecticut 06268. 674 ing between 10 and 1,000 units of this jour- nal are shown in table 1. Also shown are the average cost-the total cost divided by the number of subscriptions-and the marginal cost. The marginal cost is the ad- ditional cost of producing one more unit of output; in our example, the marginal cost is for printing and distribution and is al- ways $10. The same information is shown graphically in figure 1. Note that the aver- age cost drops rapidly as the volume of production increases; the more units pro- duced, the closer we come to the relatively low marginal cost. But the average cost never falls below the marginal cost. The importance of this attribute of the produc- tion function will become clear shortly. DEMAND Scholarly journals have two markets, li- braries and individuals. These markets differ in important ways. Libraries are generally willing to pay more than indi- viduals for journals. They are also less likely to place or cancel their subscriptions because of changes in prices, and price, al- though a consideration, is rarely the pri- mary factor in determining whether to purchase a title. Usually a library will de- cide which journals are important and purchase as many of them as it can afford. Once a subscription has been placed it is usually continued without serious review unless there is a fiscal crisis. When prices rise libraries will cancel a subscription only hesitantly, even when this means buying fewer books or making other budget cuts. For libraries, price changes have relatively little effect on purchasing behavior. 2 In the language of economics, library demand for scholarly journals is in- elastic. An often-cited textbook example of inelastic demand is kidney dialysis. Li- braries, like the kidney patient, may com- plain about prices, but they pay them nonetheless. Another important aspect of library de- mand is the limited substitutability be- tween scholarly journals. Foreign Affairs, Orbis, and Foreign Policy are all important journals covering international relations, but a library will rarely substitute one for the other solely on the basis of price. From the library's point of view, these three ti- Economics of the Scholarly Journal 675 ties are different goods. Importantly, this implies that in the library market scholarly journals are monopoly goods and pub- lishers have monopoly power. As Edward Dyl deftly points out, all publishers of copyrighted works are monopolists. 3 This view is, however, simplistic. The titles noted are distinct goods not because they are copyrighted, but because academic li- brary users, especially faculty members, who have much to say in these matters, will not generally accept an article from Orbis when what they seek is an article from Foreign Affairs. This point is critical. One can argue as Malcolm Getz has that ''the possibility of new titles by other pub- lishers entering the journal marketplace defines an upper limit on journal prices.'' 4 New scholarly journals do enter the mar- ket and in some cases they replace other scholarly journals. But it is my contention that the general case is more like the one that Getz goes on to describe: "To the ex- tent that a particular title achieves a dis- tinctive editorial position and reputation for important essays ... new entrants may pose little threat.' ' 5 Sandra Moline found a large variance in journal prices, even when the number of characters pub- lished was considered. 6 These findings, which at first may seen puzzling, are eas- ily understood when scholarly journals are seen as monopoly goods. The impor- tant point is that for scholarly journals, reputation and distinctive editorial posi- tion, once gained, are lost very slowly. It is this fact that makes them distinct goods and provides their publishers with mo- nopoly power. Popular journals are different. For ex- ample, a library might decide to subscribe to one magazine about Macintosh com- puters; it could be MacUser, MacWeek,or MACazine. Price will probably be an im- portant factor, and should the chosen title suddenly double in price, there would be little hesitancy in canceling that title and subscribing to a cheaper one. Here there is substitution, even of copyrighted mate- rial, and elastic demand. What results is a more competitive market that relies on ad- vertising to help finance publication. Individuals, at least in part because li- braries provide them with an alternative 676 College & Research Libraries Price 2SO 200 100 so \ \ \ ,_ November 1989 Average Cost Curve is AC Marginal Cost Curve is MC --------------- AC 10 1----------------- fvC so 1SO 2SO soo 1000 Quantity FIGURE 1 Cost Curves to ownership, have lower limits of the top price they are willing to pa~ for a j~umal; they are more likely to adJUSt therr pur- chases when prices change. Individual de- mand for scholarly journals is more elastic than the library demand. 7 For the sake of our example, let us as- sume that in the library market there are 250 possible subscribers and the ~ost a~y of them is willing to pay for our JOUrnal Is $250 per year. To make things easy we will assume that the demand is linear func- tion. The equation for this function will be quantity purchased by libraries = 250 - price charged libraries This is the function that describes the line ABC in figure 2. If we similarly assume that there are 750 possible individual sub- scribers and that the most any one of them would be willing to pay is $25 per year, and that once again the demand is a linear function, personal demand could be rep- resented by the function quantity purchased by individuals = 750 - (30 x price charged individuals) This is the equation for line DE in figure 2. If we combine the two demand curves to find the demand for the total market, we will have a curve that is linear but that has an elbow. Above a price of $25, the curve is the library demand curve since above $25 no individuals will purchase the jour- nal. Below $25, the library and personal demand curves must be added together. The combined demand function is if price is greater than $25, the~ quantity purchased = 250 - pnce if price is less than or equal to $25, th~n quantity purchased= 1,000- (31 x pnce) This is the equation for line ABF in figure 2. SUPPLY AND DEMAND When we bring together the supply and demand curves, we can begin to under- stand the nature of the scholarly journal . Price 250 A 200 100 50 so, 150 250 Economics of the Scholarly Journal 677 Library Demand Curve is ABC Personal Demand Curve is DE Combined Demand Curve is ABF 500 750 1000 Quantity FIGURE2 Demand Curves market. This is shown in figure 3. In a competitive market, an equilibrium is es- tablished at the point where the marginal cost curve intersects the demand curve. At this point, the cost of production will be equal to the willingness of the consumer to pay. This results in marginal cost pric- ing. The amount produced at this point is the efficient quantity and the price charged is the efficient price. Production at this point and at this price provides the greatest benefit to society. In our example, with a marginal cost of $10, the demand will be 690 subscriptions. There is, how- ever, a problem. At this point the revenue from selling 690 subscriptions will be $6,900-$10 per subscription times the 690 subscribers. The cost of producing this number of subscriptions will be $16,900- the $10,000 fixed costs and the $10 variable cost times 690. There will be a loss to the publisher of $10,000. Economists call this situation a natural monopoly. A natural monopoly occurs when the average cost falls over the entire range of demand; it is a corru."llon occur- rence where there are large up-front costs, low per-unit costs, and limited substitut- ability. This is the case for scholarly jour- nals. An important implication of a natu- ral monopoly is that, without a subsidy, no for-profit firm will undertake the publi- cation of the journal and price it at mar- ginal cost. In fact, even if a price above marginal price is charged, there will be a loss. For example, if the subscription price were $20, there would be 380 subscrip- tions, revenues of $7,600, costs of $13,800, and a loss of $6,200. A profit can be made, but only at a price where the average cost curve is above the demand curve. In our example this break-even point is a sub- scription price of $64 and 186 subscribers. This is a long way from the efficient price and the efficient quantity. The fact that scholarly journals are natu- ral monopolies explains a great deal. We see why there is a strong incentive to keep 678 College & Research Libraries the up-front costs low and to find ways to subsidize publication. To do this, editors volunteer their time and authors are not paid for their efforts. Both editors and au- thors are willing to undertake this free la- bor because, in addition to contributing to knowledge, they assume a long-term in- crease in academic reputations. The pres- sure to keep costs low may also limit the number of pages that can be published, which may in tum lead to more rigorous review and acceptance policies. Finally, if wide distribution is the goal, a way to sub- sidize publication must be found; often support comes from an academic institu- tion or scholarly society. If not, the short- fall can be made up by accepting advertis- ing. What is important to understand is that the nature of the production and the demand for scholarly journals creates this situation. Publishers' greed is not the fun- damental cause, although as we shall see later, there are opportunities for the greedy to exploit the situation. It is important to avoid another fallacy. Inelastic library demand is not all that causes high journal prices; it is the interac- tion of this demand function with the pro- duction function of the scholarly journal. Even if libraries somehow managed to change the nature of their demand so that it was more elastic, scholarly journal prices would not come down; rather, jour- nals would go out of business. Because scholarly journals are natural monopolies, they require subsidies if they are to pro- duce at the efficient quantity. 8 Without the inelastic library market, they would not be published at all. The important and diffi- cult question is how best to provide there- quired subsidy. SURPLUS AS A MEASURE OF SOCIAL BENEFIT Before going on, we need to better un- derstand natural monopolies. One might suspect that natural monopolies, which price at marginal cost, would be rare be- cause no profit can be made. Why would we want or even accept a price and pro- duction level at the efficient quantity? Why would we want a venture that al- ways operates at a loss? The answer is sim- ple: not all the benefit gained is measured November 1989 by the profit or loss of the producer. We ail understand this and have little trouble supporting, with our tax dollars, public works projects from bridges and suoways to public libraries. None of these ventures, even those that charge fees or tolls, breaks even. They all operate year after year in the red with continued public support. Even though there is no profit, society as a whole is better off when these projects are undertaken. Economists call the societal benefit in excess of cost, surplus. A closer look at this concept and its flip side, dead- weight loss, provides important insights into the inefficiencies of the current schol- arly journal system and the policy dilem- mas we face in trying to change it. It is im- portant here to keep two issues distinct. The first is the amount of surplus, and the second is who gets it. We need to consider both the size of the pie and how the pieces are divided. There are two types of surplus, pro- ducer surplus and consumer surplus. Pro- ducer surplus is the easier to understand. It is the difference between the cost of pro- ducing and the revenue received; it is the producer's economic profit or loss. 9 In our example, at marginal cost pricing the pro- ducer surplus is negative. Consumer sur- plus is less clear-cut. It is the difference be- tween the demand curve and the con- sumer's cost for the product. One way to think about this is to consider the one li- brary subscriber who is willing to pay $250 for our sample journal and assume that this $250 is reflective of the value this jour- nal has to the library. But because the price is based on marginal cost, the library is only charged $10. The difference, $240, is the measure of benefit the library receives but for which it does not pay. If we look at all the consumers who place a value on the journal above the price they are charged, we have the total consumer surplus. In our example, at marginal cost pricing, this is the shaded area in figure 4. By compar- ing figure 3 and figure 4, it should be clear that the negative producer surplus of $10,000 is offset by a larger consumer sur- plus. This surplus turns out to be $32,175. In our example, even though there is a loss to the publisher, society as a whole bene- fits. The measure of this benefit is the total Price Price 250 24.49 10 250 Economics of the Scholarly Journal 679 FIGURE3 ~ Revenue (II Loss 690 Efficient Quantity Natural Monopoly Marginal Cost Pricing AC t£ 1000 Quantity liJ Consumer Surplus FIGURE4 690 Efficient Quantity tOOO Quantity Consumer Surplus with Price at Margin Cost 680 College & Research Libraries surplus of $22,175. As it turns out, mar- ginal cost pricing creates the greatest total surplus. This is why the quantity pro- duced at this point is called the efficient quantity. This general principle explains bridges and subways. But scholarly journals are different. Public works are provided by a government. Put simply, the govern- ment's role is to collect enough money in tax dollars from those who receive the benefit, or society at large, to make up for the loss in providing the good. The gov- ernment is the transfer agent that makes it possible for a natural monopoly to operate with marginal cost pricing. To the extent that universities and scholarly societies publish scholarly journals, they can play this role. Grant-supported page charges are another subsidy mechanism. 10 But where commercial firms dominate the scholarly journal system, the operation functions largely without a transfer agent. Before we end our consideration of sur- plus, we need to introduce another concept-deadweight loss. Deadweight loss is the measure of lost societal benefit, that is, of how much the pie has shrunk. To calculate the amount of deadweight loss, we begin with the amount of surplus at marginal cost pricing and then subtract the amount of surplus under another pric- ing scheme. The result is deadweight loss. DUAL PRICING As we have seen, it is a losing proposi- tion for a private firm to produce a schol- arly journal and sell it at or near marginal cost. Therefore, it is easy to understand why publishers look for alternatives. One option in a market like that of the scholarly journal is price discrimination. Librarians usually refer to this practice as dual pric- ing. Producers can discriminate on the basis of price if they have monopoly power, if there are two or more distinct segments of the market with different demands, and if it is possible to restrict deals between the two groups. The scholarly journal market meets these conditions. There are two market segments with different demand functions and resale deals are limited. Ironically, it is libraries that create the November 1989 greatest barriers to deal making. Library procedures and the need for reliable re- ceipts make deals difficult, and operating efficiencies are usually considered more important than the savings on subscrip- tion prices. To show how price discrimination might work, consider a publisher who uses marginal pricing for individuals, charging them $10. But this publisher has decided to charge libraries $100 for a sub- scription (see figure 5). At a subscription price of $100, there will be 150 library sub- scriptions. As before, an individual sub- scription price of $10 will bring 450 per- sonal subsciptions. There will be a total of 600 subscriptions, 90 fewer than with mar- ginal cost pricing. All of the lost subcrip- tions come from the library market. The production costs are $16,000 and revenues are $19,500. The publisher earns a profit of $3,500 and substantially improves the long-term prospects of the journal. The consumer surplus is $14,625, much less than under marginal cost pricing, but the producer surplus has risen by $13,500, from negative $10,000 to $3,500. The total surplus is $18,125, and deadweight loss is $4,050 or about 18 percent. A variation of this pricing scheme is worth noting. Assume the publisher is a scholarly society with a goal of maximiz- ing the number of individual subscrip- tions and with a need only to break even financially. The price of the journal for in- dividuals could be cut in half, to $5, and if the $100 library rate was maintained, costs would just be covered. There would be 750 subscriptions: 600 to individuals, 150 more than with the $10 price, and 150 sub- scriptions to libraries. Costs would be $17,500 and revenues would be $18,000. The producer surplus would be $500. The consumer surplus would be $17,250 and the total surplus $17,750. The deadweight loss would be $4,425. This situation is in- teresting because while consumer surplus is increased, and it might be argued that society is better off because the knowledge contained in the journal is more widely distributed, the deadweight loss is greater than when the individual price is set at $10 and there are 150 fewer individual sub- scribers. Price 250 100 10 Library Market 150 250 240 • Consumer Surplus Economics of the Scholarly Journal 681 Personal Market 450 • Deadweight Loss 750 Quantity ~ Consumer Surplus Captured by the Producer FIGURES Dual Pricing If, on the other hand, the goal of the publisher is to maximize profits, prices would be set differently. A monopolist seeking to maximize profit will produce the quantity where the marginal revenue equals marginal cost. That is, where the costs of producing the last item is equal to the revenue earned on the last sale. If this is done in the two markets, the library price would be set at $135, which would bring 115 library subscriptions. The indi- vidual subscription price would be set at $22.50 and there would be 75 individual subscribers. 11 There would be a total of 190 subscriptions; cost would be $11,900 and revenues would be $17,212.50. The pub- lisher would realize an economic profit of $5,312.50, which is also the producer sur- plus. The consumer surplus would be $7,893.75 for a total surplus of $13,206.25. There would be a deadweight loss of $8,968, a full40 percent. From the point of view of the consumer and of society at large, this is the worst case we have yet considered, but it is the best for the pub- lisher. 12 The alternative pricing strategies, the quantities produced in the two mar- kets, and the profits that result are shown in table 2. Table 3 shows the surplus gen- erated in each case. If we take a broad view of the process of scholarly communication, we might find the first two examples of price discrimina- tion acceptable. Libraries subsidize indi- viduals, but the result is a widely distrib- uted journal that has the wherewithall to continue on sound financial footing. 13 In the latter case, we would probably react differently. Again libraries provide the means to make the journal successful, but this time, rather than achieving the broad distribution of scholarship, only a few can afford to subscribe. The benefit of the en- terprise goes to the publisher, not to the academic community. There are indica- tions that both strategies are used by scholarly journal publishers . Dyl' s study of business and economics journals found 682 College & Research Libraries that private publishers were more likely to discriminate and that their price differen- tials were greater than university presses' or professional associations' .14 A broader and more systematic study by Patrick Joyce and Thomas Merz found similar results across a number of disciplines. 15 THE DILEMMA What is important to note from these ex- amples is that it is the goals of the pub- lisher and the way in which the price dis- crimination is applied that should concern us, not simply the fact of price discrimina- tion. When we look at price discrimination we must be clear about what our goals are Units Produced 10 25 50 150 250 500 1,000 TABLE 1 PRODUCTION COSTS Total Avera~e Cost($) Cost() 10,100 1,010 10,250 410 10,500 210 11,500 77 12,500 50 15,000 30 20,000 20 Marginal Cost($) 10 10 10 10 10 10 10 November 1989 for the scholarly journal system before we leap to judgment. We may wish to maxi- mize total surplus or to maximize con- sumer surplus. If we hold stock in the publishing company, we may wish to maximize producer surplus. A position on the middle ground would be to maximize the amount of surplus that at the same time allows publishers to stay in business. We need also to understand that because we are dealing with a natural monopoly, and because many producers are private firms, we cannot have it all at once. The market does not have the means to make the transfers necessary both to maximize surplus and to maintain viable private publishing ventures. This is the funda- mental cause of the scholarly journal di- lemma. X-INEFFICIENCY While there are clear examples of jour- nal publishers who are in the market to maximize profits, it is equally clear that there are many who have more noble ends. But even those who are primarily concerned with the wide distribution of scholarly information are affected by mar- TABLE2 PRICING STRATEGIES, SUBSCRIPTION LEVELS, AND RESULTING PROFITS Marginal Dual Dual Dual Cost Price Price Price Price One Two Three Cost to libraries $10 $100 $100 $135 No. of library subscriptions 240 150 150 115 Revenue from libraries $2,400 $15,000 $15,000 $15,525 Cost to individuals $10 $10 $5 $22.50 No. of individual subscriptions 450 450 600 75 Revenue from individuals $4,500 $4,500 $3,000 $1,687.50 Total subscriptions 690 600 750 190 Total revenue $6,900 $19,500 $18,000 $17,212.50 Total costs $16,900 $16,000 $17,500 $11,900 Profit -$10,000 $3,500 $500 $5,312.50 TABLE 3 PRICING STRATEGIES AND RESULTING SURPLUS Marginal Dual Dual Dual Cost Price Price Price Price One Two Three Consumer surhlus $32,175 $14,625 $17,250 $7,893.75 Producer surp us -$10,000 $3,500 $500 $5,312.50 Total surplus $22,175 $18,125 $17,750 $13,206.25 Deadwei3ht loss $4,050 $4,425 $8,968.75 % of dea weight loss 18% 20% 40% Economics of the Scholarly Journal 683 ket conditions. In exploring how this is so, it is useful to employ the economic con- cept of x-inefficency. X-inefficiency results when a product is produced at greater than least possible cost. This occurs when a firm has monopoly power and in the usual example is caused by managers who give themselves unnecessary perquisites such as expensive lunches or mahogany desks in comer offices. As we have al- ready noted, scholarly journals, at least in the library market, have considerable mo- nopoly power, but as we have noted, most scholarly editors and authors work for lit- tle or no pay. Where then are the unneces- sary perquisites? We must look more closely at the goals of scholarly publish- ing. One purpose of the scholarly journal is the distribution of scholarly information and, through libraries, the creation of an archive of the cumulative knowledge of humankind. But there is a second and, to many, more important purpose. As we Price 250 • Consumer Surplus briefly noted previously, publishing is one of the primary means by which scholars achieve the recognition of their peers. This recognition in tum brings promotion, ten- ure, and financial reward. It is this latter purpose that provides the incentive for x- inefficiency. There is no question that scholars are pressirred to publish, nor is there much doubt that a sizable proportion of this publication is redundant or of less than outstanding quality. Many decry the situ- ation, but it continues and, if anything, grows worse. How can this be given refer- eeing systems and peer review? The an- swer is simple. The same inelastic demand for scholarly journals in the library market that can allow a profit-maximizing pub- lisher to reap large profits allows an editor who wishes to maximize the number of papers published in his or her journal to do so. Profits do not go to the bank; rather they are used to increase the size of the jouf!\al. The larger journal provides more 250 X-lnefficiency 120 130 1ee 250 Quantity • Partial Deadweight Loss • Deadweight Loss FIGURE6 X-Inefficiency 684 College & Research Libraries academic prestige; the editor can publish more papers by friends, former students, or simply those who share his or her aca- demic views. The result is larger, higher- priced journal with a smaller distribu- tion.16 To the established scholars who manage much of the journal system, it is the journal's role as bestower of academic recognition that is paramount, not its role as distributor of scholarly information. For the latter, the established scholar relies on other means: conferences, preprints, and a network of colleagues. To illustrate, consider the library market for our sample journal. We showed that the break-even point was at a price of $64 where 186 subscriptions would be sold. There was no producer surplus and the consumer surplus was $17,298. We will use these figures as a base point. Consider now that rather than charging $64 the edi- tor decides to print more pages, and to do so raises the price to $120. This change in turn leads to 130 subscriptions. Again there is no producer surplus and the con- sumer surplus is now $7,800. The dead- weight loss is $9,498. Figure 6 illustrates the situation. But something is not quite right. Deadweight loss implies that soci- . ety has lost, but in this case at least some of the surplus that we have counted as lost has gone to the editor and the authors. This is the darkly shaded area in figu~e 6. Not all of this benefit was lost, but because some of the articles are marginal, this is clearly not the best way for society to use its resources. The black area is different; this is deadweight loss, plain and simple, as a result of 56 libraries not purchasing the inflated journal. THE UNDERGROUND ECONOMY So far we have looked at the formal sys- tem of scholarly communication and have seen that it is a natural monopoly that pro- duces far from the efficient quantity. Soci- ety as a whole seems to be a big loser. Why then is it only librarians, and to a lesser ex- tent publishers, who are up in arms? The answer is obvious. Most scholars operate largely outside the formal market. Their transactions are underground; that is, scholars rely on personal networks, the in- November 1989 visible college, to provide much of the in- formation they need. This network has a different set of demand and production functions, and different prices are faced by the individuals in it. For scholars, sub- scriptions are only one means of adding journal articles to their personal collec- tions. Often they receive preprints or re- prints from colleagues. They can always go to their local library and make a copy of the required article, or more likely, they can send a graduate student. It is impor- tant to note that in both of these cases the prices faced by the scholar are different from those of subscribing to the journal. The price paid by the scholar is the time spent writing for the reprint or going to the library to search out the article and a few nickels for the photocopier. The only costs the scholar faces are mar- ginal costs. The up-front costs are not con- sidered. Copyright, the mechanism that is meant to provide the publisher a return on investment to cover first-copy costs, is ig- nored. Copying practices have been justi- fied by the doctrine of fair use. For the scholar, the journal system is like a public works project. Scholarly societies and uni- versity libraries finance the system much the way the government finances bridges . In doing so, they create a public good that is scholars' to use at will. Scholars work in ways that mitigate the dysfunctions of the above-ground scholarly journal market. The important question is whether the inefficiencies of the current system-the combination of high journal prices and the underground acquisition of journal arti- cles by scholars-are great enough that so- ciety would be better off providing the subsidy necessary to price the journal at marginal cost in the first place. There is no easy way to know, but clearly the costs of using university libraries can be great in both time and trouble; this part of the un- derground economy is expensive. Most studies of document availability indicate that the average user of an academic li- brary has only a 50 to 60 percent chance of finding the item for which he or she is searching. 17 Even though scholars have made adaptations that may help them in using it, the scholarly journal system as it exists today is far from efficient. ''What is important to note from these examples is that it is the goals of the publisher and the way in which the price discrimination is applied that should concern us, not simply the fact of price discrimination.'' WILL ELECTRONIC SYSTEMS HELP? Our understanding of the economics of the scholarly journal is important when we look toward the future, for there seems to be a common assumption among librar- ians that the advent of electronic scholarly communication will solve the scholarly journal problem. John Lubans, Jr., is typi- cal when he writes: In my simple thesis, electronic journals mean that libraries would no longer pay an up-front subscription cost: we would pay as we use the information in publishers' data banks. Consid- ering the cost of computer inputting and stor- age, it is unlikely that publishers would main- tain extensive back files or "inventories" ... Furthermore, publishers might even be motivated to "publish" only genuinely new in- formation and reject that which does not make an obvious contribution. 18 After contemplating his vision, how- ever, Lubans comes to the conclusion that there will be few gains. "Ultimately, elec- tronic publishing may enable us to make gains in space, but not in budgets; pub- lishers will not give up earnings regard- less of how many fewer 'pages' they rna~ 'publish' in some giant computer." 9 While the pipe dream of the great data- base in the sky and the cynical view of the forever exploited are easy answers, they both ignore the fundamental economics of the distribution of scholarship. What can we say about the economics of electronic information? To begin with, on- line systems, at least as they are currently conceived, will have a cost structure simi- lar to that of scholarly journals. The up- front fixed costs to do the editorial work and promotion will remain but are likely to decrease as authors provide copy in machine-readable forms. In addition, Economics of the Scholarly Journal 685 there are fixed costs of maintaining the distribution system, the computer, and the communication network. The variable costs associated with the production of each unit of output will likely be less than the printing and distribution of a journal issue today. At most, the cost will be a few seconds of central processing unit (CPU) time and communication costs. Like scholarly journals, online information sys- tems will be natural monopolies. The only way that a private firm will be able to exist in this environment will be to charge prices well above the efficient price and thus limit access to the few who are willing to pay a great deal. There may, however, be opportunities to improve the situation. As editorial costs fall, the amount of sub- sidy required to produce at the efficient . quantity also falls. More importantly, uni- versities already possess large computers and support communications networks. Such support constitutes an easy and ef- fective subsidy. The expectations of users of electronic systems will also change; scholars will see only the small marginal cost, not the up- front cost. They may not understand why online access to commercial information is expensive. In an electronic environment the marginal cost of distributing scholarly information may approach zero, and scholars might expect that the access to this information should be provided as a public good. They will want access to the whole universe of knowledge through their personal computers. Another predictable effect of digital in- formation will be that the underground in- formation economy will be even more widespread than it is today. As Theodor Nelson, the father of hypertext, has said: Once material goes out to the user, there is no telling what becomes of it. The user may read it on a screen, print it out or save it on a disk, and there is no reasonable way of preventing this or telling that he has done so. Thus we must live with the fact that there is no controllinJ the out- put, or its use, once it exit~ the wire. This prospect frightens publishers, and their fears may be justified. In the past, copyright has provided exclusive rights that have allowed publishers to recover 686 College & Research Libraries their up-front costs. When copying be- comes impossible to stop, copyright will be less important a protection for intellec- tual property than restricted access. But can any reasonable restriction be effec- tive? Mter all, the point is to sell informa- tion, not to horde it. The ease with which digital information of all types can be re- produced suggests that if high prices for information continue in the legitimate marketplace, an underground electronic information economy will flourish. Another scenario suggests that pub- lishers and libraries may not be necessary. A scholar-to-scholar network that com- bines electronic mail and bulletin boards may create a wired version of the invisible college. In the extreme case, each scholar becomes his or her own publisher; they pay the cost of putting their own material onto the network and they receive royal- ties directly each time their material is used. The costs of computers and com- munications will still need to be paid, but as noted, subsidy mechanisms already ex- ist. Although there is concern about refer- eeing in such a system, it is easy to imag- ine a system that would incorporate peer review. Electronic media may lower production prices and to some extent cut publishers and libraries out of the loop, but the real advantage of electronic systems will be to allow institutions to create marginal prices for their members by subsidizing informa- tion services internally. In its electronic from, scholarly publishing remains a nat- ural monopoly, and especially when com- mercial firms are th~ producers, pricing and distribution patterns will be similar to those of the scholarly journal. What changes with electronic information is that subsidy mechanisms are much easier to implement. A library, if it chooses, will be able to redistribute the information within its parent organization at the mar- ginal price. The trick for libraries has al- ways been to find the means to acquire ex- pensive information and to make it widely available to its users at a low cost in both time, trouble, and dollars. In a paper world this was a difficult, if not impos- sible, task. In an electronic world the task is much easier to imagine. It will require November 1989 the development of a technical infrastruc- ture, the negotiation of redistribution ar- rangements with publishers, and the de- velopment of internal pricing structures and accounting systems; while difficult, all of these requirements are possible. CHOICES As is often the case, understanding the economics of a problem does not provide easy answers; rather, it clarifies the alter- natives. There will still be conflicting inter- ests and different beneficiaries of different policies. But at least we have a better idea · of what the choices are and where the ben- efits fall. As we have seen, the distribution of scholarship is a natural monopoly. This means that unless there is a subsidy, the system will not work efficiently. There are then two important questions. First, is in- formation distribution important enough to justify the subsidy? And if it is, how and to whom is the subsidy provided? If we answer the first question affirmatively, we will confront political battles that can only be won if we have carefully considered our answer to the second question. Even if libraries make considerable changes in their demand for scholarly journals, which is unlikely in the short term, the production function of scholarly information will mean that the market structure and prices will remain much as they are now. The most efficient solution is to provide subsidies for the distribution of scholarship, but without fundamental changes in government information pol- icy this will not happen. In a paper system there are ways to reduce the cost of distrib- uting information within an organization. We can make libraries easier to use or pro- vide document delivery and selective dis- semination of information services. But because print is by nature cumbersome to copy, and because copyright issues are still largely unsettled, the possibilities are limited. Electronic information will provide greater opportunities for organizations to redistribute information internally. Sub- sidy mechanisms will be much easier to implement so that surplus inside the orga- nization can be maximized. The acquisi- tions of funds for subsidizing information services-the tax problem-will remain a difficult task. The argument that needs to be made is one of increasing institutional surplus through a subsidy to information infrastructure and services. The impor- tance of the natural monopoly model of Economics of the Scholarly Journal 687 scholarly communication is that it makes the benefits of such an investment clear. If we fail to make this case, we may end up perpetuating the inefficiencies of the scholarly journal. REFERENCES AND NOTES 1. For a sampling of library opinion see Richard DeGennaro, "Escalating Journal Prices: Time to Fight Back," American Libraries 8:69-74 (Feb. 1977); Richard M. Dougherty and Brenda L. Johnson, "Periodical Price Escalation: A Library Response," Library Journalll3:27-29 (May 15, 1988); John Lubans, Jr., "Scholars and Serials," American Libraries 18:180 (Mar. 1987); and Herbert S. White, "Differential Pricing," Library Journa1111:170-71 (Sept. 1, 1986). For a publisher's viewpoint see John Tagler, "Counterpoint: A Publisher's Perspective," American Libraries 19:767 (Oct. 1988). 2. The best information on library behavior in regard to journal selection and cancelation is provided in two studies by Herbert S. White: "Publishers, Libraries, and Costs of Journal Subscriptions in Times of Funding Retrenchment," Library Quarterly 46:359-77 (Oct. 1976) and "Factors in the De- cision by Individuals and Libraries to Place or Cancel Subscriptions to Scholarly and Research Journals," Library Quarterly 50:287-309 Guly 1980). See also Richard M. Dougherty and Nancy E. Barr, "Paying the Piper: ARL Libraries Respond to Skyrocketing Journal Subscription Prices," Journal of Academic Librarianship 14:4-9 (Mar. 1988); and Ann Okerson, "Of Making Many Books There Is No End: Report on Serials Prices for the Association of Research Libraries," in Report of the ARL Serials Project, Washington, D.C.: Association of Research Libraries, 1989, especially chapter two, "Description of the Problem for ARL Libraries," p.7-22. 3. Edward A. Dyl, "A Note on Price Discrimination by Academic Journals," Library Quarterly 53:161 (Apr. 1983). 4. Malcolm Getz, "How Journals Are Priced," Bottom Line: A Financial Magazine for Librarians 2:39 (Dec. 1988). 5. Ibid. 6. Sandra R. Moline, "The Influence of Subject, Publisher, Type, and Quantity Published on Journal Prices," Journal of Academic Librarianship 15:12-18 (Mar. 1989). 7. Dyl makes an interesting point in this regard. In both the library and the personal markets, it is the faculty member who makes the purchasing decisions. The difference is that in the personal market the faculty member is using his or her own money. In the library market, the money belongs to someone else. Given this situation the differences in elasticity between the two markets is under- standable. See Dyl, "A Note on Price Discrimination," p.162. White's study provides some em- pirical support for this contention. See White, "Factors in the Decision by Individuals and Li- braries," p.298-300. 8. In fact, subsidy is not the only means of providing production at an efficient quantity. Taxes or government regulations are alternatives, but because the scholarly journal market is international and because of concerns for academic freedom, these solutions are probably neither practical nor desirable. 9. It is important to note that a profit, as we normally think of it, as return of investment, is included in a firm's cost of doing business. Economic profit is profit above and beyond profit in the normal sense. 10. In an interesting study of the mineralogy and related literature, journals with high bibliometric ratings (based on citation counts) received more papers that were grant supported. These journals tended to be published by societies, to have higher circulation, and to have lower prices. Perhaps this is an indication that this form of subsidy could be effective. See Paul H. Rib be, "Assessment of Prestige and Price of Professional Publications," American Mineralogist 73:449-69 (May/June 1988). 11. Although this example is made up, it is not beyond belief. The individual (member) subscription price of Physical Review Letters in 1988 was $70; the institutional (nonmember) price was $625. For Tetrahedron the 1988 special price for individuals was DM 278; for institutions the price was DM 4,700. Sex Roles: A Journal of Research cost individuals $25 in 1988; institutions faced a subscription price of $168.50. 12. A recent study; commissioned by the Association of Research Libraries, of costs and prices of sev- eral major commercial scholarly publishing firms indicates a strong tendency toward profit- 688 College & Research Libraries November 1989 maximizing pricing and implies that the firms have considerable monopoly power. See Economic Consulting Services, Inc., ''A Study of Trends in Average Prices and Costs of Certain Serials Over Time,'' in Report of the ARL Serials Project, Washington, D. C.: Association of Research Libraries, 1989. Henry Barschall's study of physics journals also provides supporting evidence: Henry H. Barschall, "The Cost-Effectiveness of Physics Journals," Physics Today 41:56-59 Guly 1988). 13. In fact, Michael Koenig has argued that dual pricing allows subscription prices lower than would otherwise be the case. See Michael E. D. Koenig, "Serials Dual Pricing: The Librarians' Hobgob- lin," Serials Librarian 8:25-28 (Spring 1984). 14. Dyl, "A Note on Price Discrimination," see table 1: "Pricing Policies of Academic Journals," p.164. 15. Patrick Joyce and Thomas E. Merz, "Price Discrimination in Academic Journals," Library Quarterly 55:273-83 Guly 1985). See table 2: "Mean Percentage Markup, 1974" p.278, and table 3: "Mean Percentage Markup, 1984" p.279. 16. Moline's finding that categories of journals that had higher costs per thousand characters printed also printed more pages supports this theory. See Moline, "The Influence of Subject, Publisher, Type, and Quantity Published on Journal Prices," p.18. 17. See, for example, Ann C. Ciliberti and others, "Materials Availability: A Study of Academic Li- brary Performance," College & Research Libraries 48:513-27 (Nov. 1987); Rita Holt Smith and Warner Grande, "User and Library Failures in an Undergraduate Library," College & Research Li- braries 39:467-73 (Nov. 1978); or Jo Bell Whitlach and Karen Kieffer, "Service at San Jose State University: Survey of Document Availability," Journal of Academic Librarianship 4:196-99 Guly 1978). 18. Lubans, "Scholars and Serials," p.180. 19. Ibid., p.181. 20. Theodor H. Nelson, ''A Vision of the Future: Too Important to Be Left to Technicians,'' Publishers Weekly 226:51 (Nov. 23, 1984). Six Million Citations Just Met Their Match! Match wits with the BIOSIS Previews~· database. The new 1989 edition of the BIOSIS Previews Search Guide is the key to quick and effective searching of the BIOSIS Previews database. Here's what you get: • Almost 1,000 new words in the Master Index-nearly 17,000 entries. The Master Index in the BIOSIS Previews Search Guide will lead you to all the words you need to know when searching for life science literature. • The Concept Code Section has been broken down into three parts so you can more readily locate the information you need. These sections help you to match keywords and subjects with approximately 600 indexing terms. • The Biosystematic Code Section enables you to match taxonomic names with over 700 alpha-numeric codes assigned by our indexers. • New for this year is a Search System Information Section which includes vendor specific information on how to search the database. • The enhanced Searching Fundamentals Section shows you how to make your search strategies more effective and provides you with even more money- saving tips. The new BIOSIS Previews Search Guide is a perfect match for searching the BIOSIS Previews database. And this year it's better than ever. Why not start cutting down the time you spend online? Order your copy now! Just $95.00 for this invaluable reference tool. Place your order today! Contact BIOSIS, Custome~r Services Section, 2100 Arch Street, Dept. CRL1189MM, Philadelphia, PA 19103-1399 USA. • Telephone (215) 587-4800 worldwide; toll free • 1-800-523-4806 (USA except PA); Telex 831739; IIOSIS · Fax (215) 587-2016. 8/0SIS Pre•lews is a registered trademark of BIOSIS. BIOSIS is a registered trademark of Biological Abstracts Inc. ----------------------------------------------------------------------------------------~ THE WILSON INDEXES When You Need an Answer Fast and it Has to be Right Offering broad coverage, unparalleled accuracy, and unmatched currency, the Wilson Indexes are your key to important literature in business and law, science and technology, art, education, the humanities, the social sciences, and general reference. Now you can search these renowned indexes in print, online, on CD-ROM, and with patron-accessible software. B 0 0 K I N D E X E S Book Review Digest "A reputation built on many years of excellent service." - REFERENCE SERVICES REVIEW Cu111ulative Book Index Providing excerpts from, and citations to, reviews of current adult and juvenile fiction and non-fiction, Book Review Digest covers nearly 6,000 English- language books each year in the humanities, social sciences, general sciences, and library review media. Culled from more than 95 selected American, British, and Canadian periodicals, the concise critical evaluations in BRD supply: • Author • Title • Paging • Price • Publication year • Publisher • ISBN • LC number • Descriptive note • Review excerpts • Grade level • Dewey Decimal Classification • Sears subject headings for non-fiction titles· Notations of illustrations or maps. Indispensable for collection development, acquisitions, and curriculum development, BRD provides access to entries by author, or title if appropriate. A Subject and Title Index is also included. Availability __ ;...._ ___________ _ In Print: Ten monthly issues with three cumulations; annual cumulative volume. Retrospective coverage from 1905 to date; sold on the service basis (book budget). Online & CD-ROM: Coverage from 4/83. Not currently available in CD-ROM format. Compiled by professional librarians, this single-alphabet subject-author, and title index provides a complete, permanent record of the approximately 50,000 English- language books published each year around the world. Nearly half of the records in Cumulative Book Index are based on an examination of the actual book; the remainder are derived from U.S. and British MARC records and other reliable sources. Following AACR2 guidelines for descriptive cataloging and the Library of Congress Subject Headings, each main entry in CBI may include: • Author or Editor's full name • Title • Subtitle • Illustrator • Translator • Compiler • Edition • Series note • Size if other than standard • Paging • Illustrations • Binding if other than cloth • Price • Publication date • Publisher • Distributor • ISBN and LC numbers. Availability In Print: Eleven monthly issues cumulated quarterly; annual cumulative volume. Retrospective coverage from 1928 to date; sold on the service basis (book budget). Online & CD-ROM: Coverage from 1182. CD-ROM annual subscription $1 ,295. THE D To Order, or for more information H. w. WILSON on these or any of the other Wilson Indexes, simply call toll-free: COMPANY 1-800·367·6770 950 University Ave., Bronx, New York 10452 In New York State call1-800-462-6060; in Canada, call collect 1-212-588-8400. Telefax 1-212-590-1617. Fast, effective relief for · your acquisitions headaches. Unlike some CD-ROM products, which are nothing more than electronic vendors' catalogs, Books In Print Plus rM is an important acquisitions tool specifically designed to reduce the time and effort you spend on researching and ordering ti- tles. Tasks that used to keep you and your staff busy for hours can now be completed in minutes. ORDER TITLES AS EASILY AS YOU LOOK THEM UP. Custom search through the database by combin- ing any or all of up to 18 search criteria. For exam- ple, search for all titles dealing with wine published . after 1987 and a list is displayed in seconds. If you then decide that French wines should be excluded from the search, make a simple command chan~e and immediately all titles dealing with wine (w1th the exception of French wines) pub- lished after 1987 appear. The results are so incredible that once you try Books In Print Plus, it will change the way you do acquisitions forever. Exceptionally powerful, yet easy-to- use search and acquisitions software distinguishes Books In Print Plus from other CD-ROM products. Citations can be saved or printed out in seconds- in a number of formats, or sent as electronic orders to not just one, but your choice of ~ of the most important book dis- r----,.-~ TAKE TWO MINUTES AND CALL tributors.* UNIQUE WORK-SAVING SOFTWARE MAKES IT SIMPLE. . On-~creen prompts, in- stant help screens, and a special novice mode insure that you won't waste any time getting up and running with Books In Print Plus. - BOOKS IN PRINT PLUS US IN THE MORNING. To order Books In Print Plus for the IBM PC or ·Apple®Macintosh®, or for more information about how you can streamline your acquisitions process, call Bowker Elec- tronic Publishing toll-free at 1-800·521·811 0. In New York, Alaska, and Hawaii please call us collect at 212-337-6934. For tech- nical assistance call 1- 800-323-3288. You'll be feeling better in no time. ' Electronic ordering software furnished by distributors. _- B 0 W K E R E L E C T R 0 N I C P U B L I S H I N G Bowker Electronic Publishing, A division of R.R. Bowker. A Reed Publishing (USA) Company, 245 West 17th Street, New York, NY 10011 , Telex: 12-no3; FAX: 212-645-0475.