The Impact of Publisher Mergers on Journal Prices:
A Preliminary Report
by Mark J. McCabe
Assistant Professor of Economics, Georgia Institute of Technology
September, 1998

When I was asked by the Department of Justice (DOJ) to consider the potential competitive impact of a number of publishing mergers on the market for academic journals, my initial reaction was, frankly, one of skepticism. I thought to myself, hundreds of unique titles exist and the number of publishers is immense. The first step in antitrust analysis is defining the market in order to determine whether a monopolist could in fact wield power in this market. In scholarly publishing, doesn’t each unique journal title constitute a distinct market for the purposes of antitrust analysis? Certainly no one would argue that articles in Brain Research could be easily substituted for ones in the New England Journal of Medicine, much less those in the American Economic Review. If each title corresponds to an antitrust market, then owners of individual titles already have the capacity to achieve monopoly returns; a corollary is that mergers don’t matter. Furthermore, even if markets are defined somewhat more broadly, say, to include titles whose content overlaps, the likelihood that two publishers (among dozens) would together control sufficient content in enough subject areas to warrant antitrust scrutiny seemed small.

I conveyed these first impressions to my legal colleagues at the DOJ, and, well, they were not particularly enamored with this point-of-view. Their initial discussions with university librarians across the country as they began their investigations into the proposed Reed Elsevier/Wolters Kluwer merger had revealed a certain (high) level of agitation over persistent journal price inflation during the past decade or more. Furthermore, the librarians’ ire was focused on a small number of large publishing houses (large in the sense that they had large portfolios of titles) whose price "leadership" had helped decimate their budgets and imperil scholarship at their institutions.

So, what was to be done? The Antitrust Division had recently evaluated a number of publishing mergers (Thomson/West, for example). As a result of these investigations, several of my economist colleagues at DOJ had concluded that measuring book or journal content differentiation was an appropriate method for evaluating the impact of increased concentration in these types of markets; that is, only if two or more journals were close substitutes might there be any potential concern.¹ In other words, their approach paralleled my own initial impressions! Obviously, if companies were behaving anti-competitively in this market, we would have to break new ground to determine how and why.

But first, what about alternative explanations for the persistent price increases of journals? For example, perhaps the price inflation reflects increased costs. However, an extensive review of the relevant literature (published largely in the library science field) revealed that the actual costs of journal editing and printing seemed not to have suffered any unusual run-up during the past decade. Another potential and more compelling explanation relies on a simple application of supply and demand analysis. Our interviews with libraries indicate that university budgets for periodicals are relatively fixed in a given year so that "large" increases in the population of available titles might induce librarians to cancel some titles as they add new, more desirable journals to their collections (this "population" invariably increases over time). And since each issue of a journal involves certain fixed editorial costs in addition to variable printing costs, a persistent decline in a title’s circulation will eventually force firms to raise prices as they attempt to cover the fixed costs of publication. In other words, everything else equal, a smaller subscriber base necessitates higher prices. Of course, the interesting feature of this explanation is that entry by new journals is a source of price inflation.² "Demand" for new titles eventually results in higher prices across the board.

Roger Noll, Morris M. Doyle Centennial Professor of Economics at Stanford University, is an advocate of this hypothesis. To test the hypothesis, he collected price, quality, and circulation data for a sample of journals that were included in the Institute for Scientific Information indexes in a number of fields (for the period 1978-88). After estimating supply and demand equations he found that supply was downward-sloping, i.e., price and circulation were inversely related. Since entry by new journals reduces demand for existing journals, this estimated "system" predicts increased prices with entry. It should be noted that Noll’s analysis does not determine what portion of journal price inflation is due to entry (presumably this could be assessed by measuring the change in journal population during his sample period). Furthermore, his system estimation indicates that significant price inflation has occurred independent of changes in circulation. That is, even after accounting for the effect of circulation on prices, there remains a large unexplained increase in prices.

The existence of this "unexplained" residual inflation opens the door for additional and complementary theories. Over the past six or seven months, I began initial development and testing (with some assistance from a DOJ colleague, David Reitman) of a portfolio theory of journal pricing that suggests that publisher mergers of a relatively modest size can cause competitive harm. The remainder of this brief report summarizes this work.

The Demand Side
First, some useful facts regarding what the DOJ inquiry found about the demand for academic journals. Purchase of academic journals by libraries is generally based on two factors—price and expected usage. To assemble and maintain their collections, most libraries appear to construct a cost per use ratio for each title.³ Given a budget for a relevant academic field, e.g., medicine, they then proceed to rank journals from lowest to highest in that field according to this ratio, and identify a cutoff above which titles need to be canceled; conversely, if their holdings in the relevant field do not exhaust the budget, additional titles can be purchased until the budget constraint is met. From year to year, as budgets and titles’ usage change, collections are adjusted accordingly. Over the past decade or so the general trend is for increases in library budgets to lag journal price inflation; a consequence is that many library collections have suffered a fair amount of attrition.

The most interesting aspect of library journal acquisition, of course, is that individual titles within a given field are considered simultaneously. So, for example, medical libraries group titles from various sub-fields, e.g. neurology, biochemistry, clinical medicine, etc., into a single "portfolio" and broadly apply the cost per use criterion. Thus, titles compete with each other for budget dollars across an entire field, rather than across a narrow sub-field, as intuition might otherwise suggest (an intuition based on the perspective of the typical user of journal materials, and a basis for my initial skepticism). Furthermore, since journals are highly differentiated even within sub-fields, libraries try to purchase as many titles as possible (except for the most widely used journals, libraries purchase no duplicate subscriptions).

Publishers’ Pricing Strategies
Given this demand structure, how do publishing firms price their journals? First, there exist at least two types of journal companies—commercial and non-profit—that have different strategies. In general, the latter are mostly intent on disseminating knowledge, whereas the former are primarily interested in profits. Here I assume that the non-profit firms set prices to cover costs (and are thus ignored in the analysis that follows). David Reitman and I have both considered how this question might be answered for commercial firms. Although our analyses differ in certain details, they should be seen as complementary.

David’s model
David assumes that a continuous distribution of library budgets exists, i.e., libraries have unique budgets that can be ordered from smallest to largest. He shows that each journal company (owning a single journal title) will set prices so that higher quality journals will exhibit lower cost per use ratios. Given n journals, these can be ranked from lowest to highest in terms of cost per use by libraries, with the lower quality journals (that have higher cost per use) purchased by relatively high budget libraries. Conversely, higher quality journals (that have lower cost per use) are purchased by most libraries.⁴

The intuition for this particular ordering is that higher quality imparts a "cost advantage" that makes it more profitable to price low and sell widely. Given this strategy, lower quality, or "high-cost," journals find it most profitable to price high and sell to fewer, relatively high-budget libraries. Note that although the latter firms could match the "low cost" firms’ prices, this strategy is less profitable than targeting the smaller base of well-endowed customers.⁵

Mark’s model
My approach differs from David’s primarily in the assumed distribution of budgets. I consider a finite set of discrete library budget classes; each class is populated by one or more libraries. The number of journals exceeds that of budget classes (one motivation for this approach is that, in the case of medicine for example, the number of journals is much larger than the population of academic libraries in the U.S.). Given these assumptions, my model shows that single-title journal firms targeting the same budget class set prices so that their cost per use ratios are identical to each other; for higher budget classes this ratio increases. Furthermore, given some set of journals of varying quality, higher quality journals will generally target the lower budget classes, resulting in lower cost per use ratios (as in David’s approach). The intuition for these results is the same as before: higher (lower) quality implies a cost (dis) advantage, which dictates a lower (higher) price and therefore broader (narrower) circulation.⁶

While both approaches to the question of how publishing firms set their prices are based on single-title commercial journal firms, they set the stage for future analysis of larger portfolio firms.

Implications for mergers
Under David’s scenario, mergers between firms controlling two adjacent journals—i.e., journals with adjacent cost per use ratios—result in higher prices post-merger, and therefore a change in the holdings of libraries. Mergers between firms with non-adjacent journals (at least in the few simulations that David performed) appear to have no impact on prices. This is probably due to the fact that a given journal’s price is constrained by the next higher journal, in terms of cost per use. Relaxing this constraint benefits the lower cost per use journal but hurts the higher cost per use journal. Mergers of firms with non-adjacent journals are less profitable since the lower cost per use journal’s pricing is still constrained by an adjacent journal(s) that is not involved in the merger.

Under Mark’s scenario, mergers between firms that control two journals that target the same budget class or adjacent budget classes may be profitable. In other words, mergers do not have to involve journals of adjacent quality. Note that with this approach some prices increase while others decrease, depending on whether or not a journal switches budget class post-merger. For example, consider a merger between firms with two journals in the same budget class. The price of the lower quality journal of the two involved in the merger is raised enough so that these libraries would cancel the journal and replace it with one of lower quality that is now more attractive in terms of price. Using the low-cost analogy, the introduction of this relatively "high-cost" journal allows the rest of the journals in this class (including the high quality journal owned by the merged firm) to raise prices. Depending on the parameters, the corresponding increase in profits for the merged firm may or may not outweigh the losses incurred by its lower quality journal.

Much work remains to be done here, including the analysis of larger portfolio firms and the interplay between price inflation and budgets that was alluded to earlier.

Testing the Portfolio Theory
During the course of our investigation at DOJ we collected a range of data in order to test the portfolio theory of journal pricing. We collected price data for some 3000 journals, holdings information from various libraries, and quality information from the Institute for Scientific Information (ISI) (for the period 1988-98) across a number of academic fields for ISI-ranked journals. Since our progress was greatest for medical journals,⁷ my discussion is focused in that subject area. In 1998, the number of ISI-ranked medical journals published by commercial firms exceeds one thousand.

The sample period, 1988-1998, is useful in at least two respects. First, it is sufficiently long to assess whether inflation continues to plague the journal market (and dovetails nicely with Noll’s sample). Second, the period contains a number of natural experiments, i.e., publishing mergers, that help us distinguish between different "types" of portfolio effects: the effect of internal growth versus that due to acquisition. We speculate that internal growth due to entry may produce benefits (such as coverage of an otherwise uncovered field) that help offset any intentional competitive harm; any harm associated with acquisitions, on the other hand, is less likely to be balanced by substantial benefits—titles are simply reshuffled, and any fixed cost savings seem to be small.⁸

We adopted a reduced form or hedonic method of estimation, where prices are regressed on several factors: firm portfolio size, journal quality (measured by ISI citation level), a Brandon-Hill dummy (the Brandon-Hill journal list⁹ indicates essentially whether a journal is a general or specialized title, and is thus an exogenous measure of circulation "potential"), and a number of variables that capture the effects of residual price inflation and idiosyncratic firm-specific effects. For further technical details see the short empirical appendix that follows.

Briefly, our results for journals sold by commercial publishers indicate that prices are indeed positively related to firm portfolio size, and that mergers result in significant price increases. For one specific transaction we investigated—Wolters Kluwer and Waverly—the results predicted an average price increase of between 20-30% for the affected medical titles. What is particularly noteworthy are the modest portfolio sizes involved. Based on the total number of ISI journals sold by commercial publishers (with portfolios of ten titles or greater), Wolters Kluwer and Waverly have shares equal to about 11% and 5%, respectively. Contrary to the current rule of thumb in antitrust—the combined shares of the merging firms typically need to approach 35-40% before measurable effects arise—these shares are relatively modest.¹⁰ An explanation for this anomalous result is that unlike most markets, where buyers purchase a product from one of several sellers, in the market for journals, buyers purchase titles from as many sellers as possible. This strategy delivers more market power to "small" firms, hence the large impact from modest-sized mergers.

Finally, even after controlling for the effects of firm size and other variables, there remains significant residual price inflation. One explanation is that our portfolio approach and Noll’s entry story are complementary—both factors contribute to the observed price inflation separately.¹¹ And the residual price inflation in each estimation reflects the impact of the omitted factor, whether it be firm portfolio size or declining circulation. Clearly, one important avenue to pursue in the future is how these two factors should be incorporated into a single estimation (the medical libraries holdings data should be useful in this regard since it provides snapshots over time of how journal circulation changed for a representative sample of libraries).

Implications and Future Directions
When the proposed merger between Reed Elsevier and Wolters Kluwer foundered earlier this year, opposition from antitrust authorities in Europe and the U.S. was cited as a primary cause. Although no formal complaints were filed by agencies on either side of the Atlantic, regulators had sent a variety of signals indicating their serious concerns. Negotiations with the European Union had progressed the farthest and it appeared that the proposed deal would proceed only if the parties agreed to significant divestitures. It was widely reported at the time that the EU’s preferred set of divestitures upset the financial logic of the merger and resulted in its demise.

What is interesting here is that the EU’s main focus was not on academic journals, but rather legal publishing (in Europe), and that its theory of anti-competitive harm was based on traditional antitrust principles, i.e., excessive overlap in content (and therefore similar to the DOJ’s approach to the Thomson/West merger).

The U.S. focus, of course, was far different, in part because European legal publishing was not germane and because the model of harm relied upon was novel. Though one can only speculate on how a U.S. antitrust case might have proceeded, it is clear that the combined Reed Elsevier/Wolters Kluwer entity would have controlled large journal portfolios in a number of broad fields, including biomedicine. Assuming that these broad fields constituted antitrust markets, some of these portfolios would have crossed the U.S. Government’s threshold with shares in excess of 35%. Based on the preliminary results discussed here, such a merger would have resulted in substantial price increases over time. If the U.S. had filed a complaint and had been successful with this market definition, an important legal precedent would have been set, one that would have made it easier to employ a portfolio theory in mergers involving combined market shares less than the threshold and/or a large firm buying a relatively small portfolio of journals (see footnote 10). Unfortunately, since any future deals are likely to be relatively small in scope, opposition to journal mergers will need to be successful in both dimensions: market definition and market shares.¹²

This increased burden is why further refinement of the work started over the last eight months is necessary. Using the price, portfolio and holdings data collected during the course of our investigation, I hope to further develop our portfolio theory and to test the robustness of our initial empirical results. Although antitrust policies in the U.S. and Europe have changed considerably over the past two decades in response to a better understanding of market dynamics, this type of reform requires the development of new and persuasive evidence that existing policies are inadequate.

Empirical Appendix
Hedonic estimation does not provide estimates of the "structural parameters" from a supply/demand system. This can limit the quality of inference. For example, suppose one regressed auto prices on a set of car characteristics, e.g., horsepower, passenger capacity, transmission type, etc. If we found that car prices were positively associated with horsepower, should we conclude that the cost of additional horsepower is positive or that demand for larger horsepower engines is greater? In the present case, this inference problem is not likely to affect the firm portfolio size parameter. Since a merger should not increase costs, a merger-related price increase in the hedonic framework is demand-related, i.e., a result of enhanced market power.

Another concern whenever "firm size" is specified as an explanatory factor in a price or profits equation is the issue of endogeneity. What determines firm size? If firm size is merely a proxy for some other X-factor (say, lower costs, or skill at introducing new journals) it is possible that any conclusions regarding the negative effects of firm size per se might change if this additional factor(s) was included in the analysis. This problem is most acute with cross-sectional data since firm fixed effects are omitted and no changes in the relevant variables can be observed over time (see the Handbook of Industrial Organization chapter by Richard Schmalensee on this topic¹³). However, panel data (cross-sectional data over time) of the sort collected here is ideally suited to address these endogeneity concerns. A panel allows us to observe the impact of changes in portfolio size over time holding firm attributes constant. And to address any lingering concerns that these changes in firm size are masking some X-factor (as noted above), the data allows us to distinguish between increases in firm size due to internal growth versus growth from merger. Since we are able to control for quality changes, a significant (in the statistical sense) post-merger price increase is a good indication of an anti-competitive effect.

Copyright © by Mark J. McCabe. The author grants blanket permission to reprint this article for educational use as long as the author and source are acknowledged. For commercial use, a reprint request should be sent to the author <mark.mccabe@econ.gatech.edu>.