The death of the subscription model for academic materials has been loudly proclaimed for some time now. There are many reasons for this, but the principal one is that the current financial situation at many libraries has made the cancellation of subscriptions necessary. Libraries are not growing, but they have to contend with more publications being presented to them every year (the growth in the number of publications will never stop) and price increases for publications that they already subscribe to.  Add this together and it immediately becomes apparent that libraries would be squeezed even if their budgets were not reduced. This cannot go on forever. So the subscription model must go — it is, to use that famous word, unsustainable.

I understand the plight of libraries, but I am doubtful that the subscription model is going away anytime soon, at least for journals. This does not mean that is is not changing. It is changing, in its aspects and proportion. But the economic forces driving the subscription model are too strong to disappear. In any event, extrapolating far into the future on the basis of the current weak economy leads one to gloomy conclusions — too gloomy, says this congenital optimist. I encourage everyone to check the balances in their 401(k) and 403(b) accounts and compare the sums to those from March 2009. The increase (only a terrible investor could have lost money in that time frame) is being more than matched by the growth in university endowments, at least at those institutions fortunate enough to have an endowment. We may see some library funding restored by 2013. Cheer up, everybody, and subscribe.

Over the past several months I have been collecting anecdotal information about subscriptions and also had the opportunity to participate in a lively thread on Ann Okerson’s Liblicense mailgroup. I have been surprised by what I have found. For example, more libraries appear to be disengaging from “Big Deals” than I had supposed, and there is growing interest in pay-per-view.  One reason that the extent of Big Deal cancellations is not better known is that publishers are loath to let this information out, fearing a domino effect. The most surprising thing that I stumbled upon, however, is that not everyone defines the subscription model and the Big Deal in the same way.

Concerning the Big Deal, for example, some people think of it as an enormous aggregation offered by one of the largest 4-5 journals publishers; others think of it as any aggregation from any publisher big enough to have enough journals to aggregate. Some librarians like Big Deals, others do not. (To listen to the rhetoric surrounding the Big Deal, you would think that this marketing practice was the equivalent of an occupying army.)

By the “subscription model,” some people mean only arrangements for very large aggregations — that is, Big Deals. Others (I am in this camp) think of it as a form of service that is applicable at any scale. Another point of confusion is whether subscriptions apply only to library purchases or to individual purchases as well (I say yes to the individual option).

It’s not a bad idea to know what we are talking about when we refer to the presumably doomed subscription model. From a business point of view, subscriptions are contrasted with single-unit sales. A subscription involves an ongoing payment for an ongoing service. Accounting rules capture this distinction nicely. So, for example, if a publisher sells a book for $20, the figure of $20 is recognized as revenue in the publisher’s ledger. But if a publisher sells a one-year subscription to a journal for $480, then the publisher only recognizes one-twelfth ($40) of that income each month.

Now, accounting is a black art, but the business folks reading this, whether they work for for-profits or not-for-profits, whether for publishers, libraries, or anything else, know that cashflow and income are not the same thing. When you sell things as single units (e.g., a book), your cashflow trails behind your recognized income because you are typically not paid until after you ship the book. (This changes a bit with e-books.) But when you are paid up front for a journal, you have the use of the cash for months ahead of time. This is why journals publishing is such a good business: it throws off great amounts of cash, and it does so early.

I go into these arcana of accounting to make the point that the subscription model is a business model, and as such, it concerns itself with dollars and how they are made. No journals publisher currently enjoying the economic benefits of the subscription model will willingly switch to a single-unit model or selling on demand.  Subscription marketers will only engage in single-unit sales (e.g., PPV) when they are forecast to yield incremental revenue — that is, revenue on top of the already existing subscriptions.  Typically this means that an established merchant of subscriptions will adopt a single-unit program only when it is designed to reach a new market segment — for example, libraries that are not subscribers or a scientist working in the research lab of a corporation.

Librarians who are demanding that publishers that currently sell subscriptions switch to reasonably-priced PPV will have to wait a long time.

Of course, you can imagine someone who does not have an established cash-rich subscription business to protect getting into this market with attractive single-unit pricing. Buy from me and save! As an introductory offer, this is not a bad idea, but over time certain attributes of this market will emerge. First, our disruptive publisher will soon realize that if you can switch customers to subscriptions, you can significantly improve cashflow on the same level of sales. Second, and more importantly, people don’t buy content of any kind based on price. They buy it for quality, which is not determined absolutely but subjectively–but just because it is subjective, that does not mean that it does not exist. This is an editorial game, not a commodity business. Please don’t take my word for this; consult your own behavior when you are buying books for yourself. You can get the new biography of Montaigne for $25 or a paperback technothriller by William Dietz for $7.99. The difference in price is $13.01, but while a student of Montaigne may also have a weakness for Dietz (doubtful), no one would think that the latter is a substitute for the former.

The Internet does not commoditize content. The Internet commoditizes bad content. This is not an age of content abundance. High-quality content, by definition, is always scarce.

It’s not just superior cashflow that makes subscriptions such an attractive business; it’s also what is known in the trade as the reduced cost of customer acquisition. A publisher may have a hard time to get someone or a library to purchase a subscription, but once that order is placed, most subscriptions, except for content of poor quality, are renewed year after year. This means that a single act of salesmanship results in an ongoing revenue stream. For this reason, publishers are prepared to invest heavily to get that initial subscription and often lose money for the first year or two (in some instances as many as five years) of that subscription.

When you move to single-unit sales, on the other hand, the cost of marketing goes way up. This is because all the marketing expenditures must be placed against the sale of a single unit — a very expensive proposition. In fact, the current pricing of PPV, high as it might seem, is actually being subsidized by the income from subscriptions. It is for this reason that the single-unit book business is a bad business and the subscription journals business is a good business. This also explains why so many more academic publishers are now attempting to sell libraries aggregations of ebooks: they are trying to switch from the crummy economics of single-unit sales to the rich economics of subscriptions.

So rather than talk about the death of the subscription model, we should be talking about its progressive growth and its penetration into new content types and markets.

Someone who is not a fan of the subscription model, especially a librarian who is wrestling with reduced budgets and the inflexibility of many subscription programs, will find this explanation of how subscriptions work to be wholly unsatisfying. They would be right. The subscription model is subject to abuse, and where profit is allowed and aggressive trading practices stay inside the law, we will have abuse. Those of us who subscribe to cable TV at home see this with every month’s statement. The economics of cable provide the best value when customers subscribe to a package:  so much programming for one fixed cost. The cable companies argue, correctly, that if you had to purchase each program or channel one at a time, the cost would skyrocket (naturally, they don’t give you the choice). The abuse comes in when the size of those cable packages continues to grow, thereby increasing the price of a monthly subscription.  It’s still a great value when you think about it as the cost per program, but the absolute price has risen sharply and it has done so by adding less desirable programming to the package. How I long for a cable service that does not require me to pay for ESPN, Fox News, and the Home Shopping Network! The cable TV business model, in other words, is the Big Deal for the home video market. How many consumers like that?

If there were a large-scale attempt to unbundle subscriptions, publishers would have to wrestle with different economics. In effect, from a business point of view, they would have to turn the journals business into the book business. They would begin a process of price increases that would make single-article sales terribly expensive. Such prices would deter purchases, which would result in increased unauthorized content-sharing — which in turn could lead to copyright litigation of the kind we have seen in the music business. Publishers would also go to great lengths to make sure that every single article they published would find a market of sufficient size to make that one instance of publishing profitable (allowing for subsidies among not-for-profit academic publishers). This is precisely how the book business works and the reason that so many university presses are tottering.  And let’s not stop there:  let’s add patron-driven acquisitions to the mix, which will result in even lower unit sales for all but the best titles and delayed payments even for these. Moving to on-demand purchasing for journals articles would not so much unbundle subscription-based scholarly communications as it would unravel it.

Such an attempt at unbundling would initially be a boon for various open access services, which would fill the vacuum created by the collapsing subscription model. But even here the superior economics of subscriptions is likely to emerge in time. This is already happening at BioMed Central and PLoS, where in addition to fees paid by authors to publish, revenue is derived from so-called institutional sponsorships, which are subscriptions by another name. But it won’t stop there. The competition for these services is just now beginning to heat up as more and more author-pays open access services (Sage Open, etc.) come on stream. This will put downward pressure on how much such services can charge authors.  Such organizations will in turn look for ways to stabilize their revenues, and one way to do that will be to invite authors to become “members” — that is, subscribers — with the ability to publish multiple articles over time as a membership benefit. Peer into the future a bit and PLoS begins to look like Comcast or Time Warner Cable.

This does not for a minute mean that libraries should not challenge the subscription model. Quite the opposite; in fact, I would argue strenuously that libraries can and should be much more aggressive in their dealings with vendors. The question is not whether libraries should be trying to reduce their materials costs; of course they should; the question is what is the overarching strategy and what is the desired outcome. The problem is not with the subscription model — which does in fact reduce costs, just as publishers say; the problem is with its abuse — which does in fact increase costs, whatever publishers say.