Watching your cards in The Big Deal Ted Bergstrom, Turin Conference, May 31, 2012

The Librarians’ Shopping Problems Problem 1: Delegation Like physicians choosing drugs, and college professors selecting textbooks, librarians make choices for consumers who aren’t spending their own money. – Physicians choose treatments for patients and insurance companies pay. – Professors choose texts for college students and parents pay. – Librarians choose journals for professors and “the university” pays.

Problem 2: Unreliable signals Faculty arguments for purchases are fervid, but not always entirely credible.

Problem 3) Complexity The major publishers have contrived to offer all-or-nothing deals of mind- boggling complexity. – Elsevier packages more than 2,000 journals. Springer 1,900. Wiley 1400 – Packages contain journals in more than 100 distinct disciplines. Who can say what the package is worth?

Problem 4) Monopoly Probs 1-3, delegation, unreliable signalling, and complexity lead to price inelastic demand. Markets with price inelastic demand are a monopolists’ paradise. Monopoly

Other Examples Prescription drug industry, monopoly sustained by patents, delegated purchases, complex evaluation College textbooks. Consolidation to three major publishers, delegated purchases, inelastic demand. (College degree offers huge consumer surplus. Publishers grab a bit.) California electrical power. The state agreed to guarantee a price ceiling on electricity. The state would buy electricity at whatever it cost and sell to Californians at a price no higher than the ceiling. Enter ENRON.

What do they cost? For some years, Preston McAfee and I have maintained a website that displays and compares institutional subscription prices per citation. We now list more than 9000 journals. We find dramatic differences between the prices of journals owned by large commercial publishers and those owned by nonprofits It has been argued that these are misleading because they don’t account for large discounts for package deals from big commercial publishers.

Bundle price project We and Paul Courant of Michigan library decided to collect prices paid by major universities for bundled contracts. Problems: – Confidentiality clauses. – Complex terms Response – State Freedom of Information Act Requests – Collect Entire Contracts – Cooperation and Encouragement from ARL

Hell’s Grocery Store: A Fable for Librarians

Imagine a grocery store in a land of protected monopolies: The only seller of much-desired goods. Overpriced, but there’s nowhere else to go for these items. Shoppers chose items they wanted and paid posted prices. Some bought many items, some bought few. All items sold at well above cost

How to sell more stuff at prices above cost: Grocer’s Big Idea Double all prices, but top up everybody’s cart with All the things they chose not to buy. Shoppers pay twice what they used to pay for the stuff they used to buy, but get full carts. – They could still buy items singly but they cost twice as much as before Shoppers who bought more stuff in the past pay more for the same amount of stuff as those who bought less

Shoppers complained! Grocer replied: “You have nothing to complain about. Now you get a cart full of groceries every week. Some of you used to get only half a cartful or less. You pay twice the price, but you are getting more than twice as many groceries. We have cut prices, not raised them.”

Shoppers worried How much of a discount am I getting, when my shopping cart is topped up with things I didn’t choose to buy at regular price?

Sound familiar? Commercial publishers pulled off this trick. Only a little more gradually. Price increases of about 7% per year after Big Deal initiated. Doubles prices in a decade.

So What’s in the Contracts?

Big Deal Discounts Elsevier’s Freedom package includes almost all of their journals. Purchased one-by-one, 2009 total cost is about $3.1 million Example: U of Michigan paid $2.2 million for its Freedom Package That’s a 30% discount, right?

Not Exactly If Michigan had spent its $2.2 million with Elsevier on single subscription journals, it could have obtained journals that get 91% of all the citations to Elsevier journals. So, for Michigan, the Big Deal Price is really only a 9% discount from list

2009 Payment Apparent Discount Actual Discount Illinois$2,300,00025%6% Michigan$2,165,00030%9% Maryland$1,760,00043%15% Wisconsin1,215,00061%28% Iowa$1,421,00054%22% Idaho$751,00076%44% Wyoming$497,00084%54% Elsevier Freedom Package Big Deals 2009

So, how good are these deals? Let us compare 2009 prices paid by large research universities per ISI citation and per article. – Elsevier’s Freedom package – Packages offered by major professional societies.

BundlePer citePer article Elsevier (U Mich)$3.60$15.16 Am Biochem Soc$0.20$0.95 Am Physical Soc$0.45$1.10 Am Soc Microbiology$0.45$1.20 Oxford U Press (Colorado)$0.55$2.15 Am Chemical Soc (U Mich)$0.65$2.85 Am Geophysical U$0.90$2.65 IEEE$1.05$2.25 Am Medical Assoc$1.05$5.90 Bundle Prices Charged to Large Research Universities

BundlePer citePer article Am Soc Biochem18/115/1 Am Physiological Soc10/1 Am Physical Soc8/113/1 Am Soc Microbiology8/112/1 Oxford U Press (Colorado)6.5/1 Am Chemical Soc (U Mich)5.5/15/1 Am Geophysical U4/15/1 IEEE3.5/16/1 Am Medical Assoc3.5/16/1 Ratio of Bundle Prices: Elsevier to Societies

Bargaining in Hell’s Grocery Commercial publishers maintain the fiction that contract prices are rigidly tied to current prices of the stuff that was in your cart when the Big Deal began. Some libraries seem to believe this. Evidence suggests wide variation.

Variation in Elsevier Contracts UniversityEnrollment2009 Price Texas47,000$1,500,000 Georgia33,000$1,800,000 Michigan39,500$2,200,000 Wisconsin35,000$1,200,000 Colorado28,000$1,700,000 Kentucky23,000$1,300,000 Cal (scaled)27,000$1,100,000

So, what to do? Consider dropping overpriced Big Deals Go a la carte with best deals from big publishers Works for Stanford and for Cal Tech Alternatively, bargain hard on big deals, keeping a la carte option in mind. Maintain site licenses only for publishers that price close to average cost.

Who needs libraries any more? A more radical idea… Ruins of Library at EphesusRuins of Stanford Library (1906)

Libraries and Journals When people shared paper copies of journal articles, they were needed for physical storage and a central sharing place. When paper journals cost real resources, librarians were needed to decide which were worth having in a shared environment. With electronic distribution, physical library plays no role. If marginal cost is zero, libraries act only as toll collectors.

Economic prescription? If you want to allocate resources efficiently, use the price system. Let users pay for what they get. They will economize if its their own money. Complexity is not a problem, since nobody makes a central decision about who should have what. Commercial publishers already offer pay-per-view. Let subscribers use it.

Is this the right advice? On target for California power. Less obvious for medical treatment and textbooks, since consumers may not be as well-informed as experts who decide for them. How about academic researchers? Who is better able to decide values of article access than they? – Librarians and university budget officers? – Maybe not…

An economic case for subsidizing academic journals Technology: Journal publishers have fixed costs. Recruiting editors, harassing referees, typesetting, copy-editing, setting up server. Marginal cost of allowing acess to one more user is almost zero. Efficient outcome is either to allow user access at zero cost Or not to publish the journal at all if setup costs exceed value to users.

Competitive prediction If journal supply were truly competitive and users paid the subscription costs, free entry and competition would drive prices down to just cover average costs. Journal supply is not competitive. Copyright law and coordination costs assure this. Even with no library subscriptions, there would be profits in journal publishing. But demand would be much more price-elastic and prices would be lower.

Limited role for central purchase Most (but not all) non-profit professional society journals sell at a small mark-up over average cost. Median price per article of for-profits in most fields is 3 to 4 times that of non-profits. Libraries could subscribe to journals that cost no more than, say, 1.5 times as much as average non- profit and let their patrons access these journals for free. This allows efficiencies of marginal-cost pricing.

What would happen to commercial publishing? Currently, commercial publishers charge about $35 per article for downloads. If informed users spend their own money for downloads, demand will become much more price elastic. When demand becomes price elastic, monopolists cut their prices. Double whammy. Authors would know that overpriced journals will have fewer readers. Will prefer publishing in cheaper journals.

That’s all for now

What happens when 5-year Big Deal contract expires? Faculty addicted to online access. Must negotiate new contract.