About a decade ago, we experienced the excitement of contemplating cheaper online journals and free music. The exuberance surrounding these visions of a market-free utopia swept up many, and some dwell on the potential still. The early days of e-books held the same promise for a short time. Yet, e-goods are returning to the price levels they once were thought to have abandoned. Music is firmly back to its pre-Napster price levels — the new Coldplay album costs $9.99. E-books are headed in the same direction — the new biography of Steve Jobs is $16.99 for the Kindle version. More intensely niche, scholarly materials never flagged much throughout the tumult, and continue to gradually increase in price as the years go by despite the majority of access coming through online channels.
We all know now that providing digital information isn’t a great way to save money — variable expenses are lower but fixed expenses are far greater, with technology, technical staff, creative staff, and editorial staff all adding to the costs of creating digital products. And digital goods are becoming increasingly valuable in a digital ecosystem — the value of having a PDF was made higher by email, large hard drives, and local file systems. The value of a virtual pig in Farmville is high because the game makes you competitive, and competition has a perceived value.
Price inelasticity is the underlying reason prices have clawed their way back to higher levels. Prices for unique, narrowly appealing materials are inelastic — there’s no benefit to lower prices because doing so won’t increase sales enough to justify the lost revenues. A $0.29 article on an obscure scientific finding or refined medical procedure isn’t likely to shoot to the top of the bestseller list based on price. Substitutability in inelastic markets is low — while some claim Coldplay is just “U2 light,” the two groups really can’t be substituted — so prices are inelastic. There is only one Lady Gaga, only one Walter Isaacson book about Steve Jobs, and so forth. Scholarly materials are price inelastic, as are prices for books by famous authors, music by famous bands, and so forth. Prices find the natural equilibrium between market size and profitability, and they generally stay there. And prices tend to be higher the smaller and more affluent the market.
Open access publishing is a new model for value, but one that has an interesting twist to it — there is a commodity at its heart. That is, “getting published” is the value proposition, something that PLoS, BioMed Central, Nature Communications, BMJ Open, and others can provide. Because it’s a service that is largely undifferentiated, the commodity nature of the service dominates.
You can see it a bit in the pricing. The average price for publication is about $1,900, with $1,350 the most common amount, and the standard deviation being about $1,100 (with two standard deviations in the small sample of prices I looked at). This suggests a tight distribution around the mean.
The more publishers who provide the OA publishing service, the greater the opportunity for price competition. Differentiating on quality of content is difficult, because every author paying for the service thinks their work is very good. There is no way to charge someone more by stating that their content is marginal and therefore requires a higher fee, just as there no way to charge someone less because their content is truly very good. Instead, OA publishers have to compete on venue quality, and that gets to branding and publisher reputation. This immediately raises the question of what an OA publishing itself does for overall venue quality. Chances are, it does little to improve it, and is at best neutral. So the competition around venue quality is also a bit hamstrung.
To me, OA publishing looks much more like a commodity, cost-based market than a well-differentiated value-based market. Currently, the variability is relatively slight in the services and branding provided; hence, the general and so far rather enduring grouping around the mean as pricing goes. With mega-journals like PLoS ONE sweeping up a broad range of topics — from ecology to women’s health — the median price for OA publishing isn’t likely to rise any time soon. In fact, I believe OA prices will tend downward as competition for a large but finite set of papers increases.
Publishing for an audience allows a business to find a balance between audience size and value, and price inelasticity tends to push prices upward over time. Publishing as a commodity introduces elasticity — the lower the price, the more authors you attract, and the better your business. With that working in their favor, prices for OA publishing should start trending down as the number of outlets increase. And, with that mechanism at its heart, high-volume OA publishing seems structurally inescapable.
Discussion
22 Thoughts on "Content as Commodity — Price Elasticity and New Business Models"
“And, with that mechanism at its heart, high-volume OA publishing seems structurally inescapable.”
It would be great to have a follow-up article where you share your view on the implications of this.
Kent, while I accept your argument about price (in)elasticity, I think it needs to be noted that the economic model gets complicated when we realize the following:
1) Price sensitivity changes when authors don’t spend their own money. For PLoS ONE, I can imagine that some authors may be sending the receipt to their institution, library or granting organization; however, some may be paying out of pocket or discretionary funds. On the other hand, I don’t imagine that many (if any) authors of Cell articles are paying out of pocket, which is perhaps why the journal can charge $5,000.
2) The willingness and ability to pay may depend on the service one receives from the publisher as well as the perceived “brand” of that journal. Two journals offering exactly the same publishing service may be able to charge different APCs if they have radically different impact factors, for instance. Conversely, a journal may not invest in great author services (copy-editing, for example) if it drives up the price of the service.
3) Scholarly publishing serves not one market, but many markets, each with its down unique attributes. For example, Sage Open is still charging its introductory price of $195 (discounted from its regular fee of $695), in part because the market for social sciences and humanities publishing won’t support the same fee structure as in the biomedical sciences.
Even spending someone else’s money doesn’t mean that prices won’t be known, compared, and so forth. In fact, it may increase scrutiny as OA continues to grow.
As for competitive attributes, you’re right, there could be differentiation on that front, but we’ll see. I have a feeling that the bulk of the model will continue to be high-volume, which will drive down prices overall, or at least keep them from increasing.
And, yes, each market will have its price point.
Differentiation by offering a higher level of service seems one way to set oneself apart from the seemingly identical hordes of “everything bucket” journals that are currently being launched. Doing this seems to inherently favor the big publishers over the smaller ones because of economies of scale. A large publishing house can offer a higher level of service for a lower price than a smaller one because the scale of their operation creates lower costs for those services. It’s another reason why this style of journal may just further solidify the hold that the small number of enormous commercial publishers have on the market.
I agree with all three of Phil’s points. Theoretically, OUP could charge a higher author fee for publishing an OA book than a smaller, less prestigious press. It will be interesting to see if this happens, when we get farther into OA book publishing and its financing.
…variable expenses are lower but fixed expenses are far greater, with technology, technical staff, creative staff, and editorial staff all adding to the costs of creating digital products.
Could you provide a citation here? I have no doubt that this is to some degree true, but I would be curious if it was possible to quantify exactly what that difference in fixed costs is — after all, there are a lot of fixed costs for a print publisher as well.
I wish I could, but I think we’re still in the stage of empirical observation, hampered to some extent by the proprietary nature of business expenses. The evidence is all around us, especially the recent news that Wikileaks may be bankrupt soon, despite having few editorial processes, content “supplied” to it, etc. Every publisher I know of is experiencing this — hiring more technology, management, marketing, advertising, and development staff to deal with the online world while printing costs decline as print circulations dry up.
Anyone have that elusive citation? Or anyone pursuing this research now?
see:
King DW. 2007. The cost of journal publishing: a literature review and commentary. Learned Publishing 20: 85-106. http://dx.doi.org/10.1087/174148507X183551
I can tell you from my experience as a publisher of scholarly books that the variable costs for paper, printing, binding, and warehousing are roughly 25% of the overall cost of publishing a book, and the new fixed costs associated with publishing electronically are about on the same order of magnitude, depending on how fancy you want to get (instituting XML workflow, for instance, would add cost). I would venture an educated guess (since I also published journals at Penn State) that this is true for journals as well.
Am I wrong in assuming that editorial staff are not unique to digital publishing? I mean, you’ve got that with print or online so I wonder if it’s fair to say that this is one fixed expense that is “far greater with technology.”
Also, technology in terms of hardware and software closely resembles a commodity where prices tend to fall over time.
I would even go further and say that technical staff also tends to act like a commodity whereas new skills become commonplace very quickly in the IT world and therefore the price you can charge for such skills tends to fall.
BTW, I agree with the bulk of the post, but just thought I would see if the assertions about fixed expenses assertions are valid or not.
Great post. You’ve made the argument that many OA advocates make – that non-substitutability of content fosters higher prices (what most such commentators mean when they refer to a monopoly, although you’ve characterized this more accurately as price in elasticity) while publication-side fees promise a more cost-based market (although Phil offers a useful caution that protecting authors from these fees may undermine the market effect). Isn’t this what Monbiot, for all his hyperbole and emotional rhetoric, was getting at?
I think my implicit argument is a little more nuanced. I hope so, at least. The business observation is that right now, OA publishers are peddling something that has little differentiation and therefore can’t command a price sufficient to withstand a price war. To authors, they may all appear roughly equivalent. As other comments have noted, differentiation may be possible, but it will take some effort and brand expenditure. What will be a good differentiator? Will brand be helped or hurt by the OA affiliation as prices sort themselves out?
It is nice to have a calm discussion about these things. Monbiot’s hyperbole and emotional rhetoric makes it hard to know what he was getting at, for me at least.
One thing that seems certain is that the present situation is unstable. What is important here is that the economics are truly strange. The original producers (authors) have a choice between giving away their product and paying someone to take it. Ironically it sounds a little like a waste disposal problem, but in no case is it like a commodity.
It strikes me that with author pays we are really talking about the rise of a new industry. That is, subscription publishing is one industry and OA is another, but they use the same inputs. They are different industries because different customers are paying for different products and/or services. Subscribers are paying for a product but authors ware paying for a service. These differences are so large that it is very hard to imagine what the ultimate equilibrium condition might be.
Part of the problem is that the OA industry is still so small that the instabilities are not noticeable. As far as I know subscription journals are still getting enough papers to fill their volumes. But imagine OA getting 50% of all submissions. Conversely, if an OA giant emerged and published 100,000 minimal review articles a year would tenure committees still consider that being published?
Given these large instabilities, speculation is fun but forecasting is not appropriate.
I quite agree: we live in interesting times, don’t we? I’m just glad I have no personal stake in the outcome, as someone semi-retired.
Indeed, this may be a time of great gains and losses. Opportunity is always risky.
At first, it seemed to me that Kent’s statement about the increased fixed costs of conducting both a print and electronic business was true to the point of being self-evident. I also agree with Kent that these are early days and due to the nature of the publishing industry, getting real data is difficult.
However, there are a number of publicly traded publishers on the stock market. I was unable to find cost of goods numbers for either Elsevier or Wolters Kluwer (they are out there, I just need to search more thoroughly), however, the good people at Wiley make their numbers readily available.
Wiley’s COGS:
2007 28.6%
2008 27.5%
2009 27.2%
2010 26.3%
2011 25.2%
During this time period Wiley’s sales have increased by 42.9%.
What does this tell us? If Kent is right; (fixed costs are growing as a percentage of all costs), one would expect to see total COGS (which includes both fixed and variable inputs) grow at a slower rate than sales (since fixed costs are not impacted by sales levels). Indeed, that is exactly what you see here. However, there is a problem in that (I suspect) many of the fixed costs for electronic delivery end up in the general management and administrative side of the income statement.
So here are the G&A expenses (including depreciation and amortization)
2007 56.2%
2008 56.8%
2009 57.7%
2010 58.1%
2011 57.2%
What to make of this? There is no real general trend here. But if Kent is right (fixed costs are growing as a percent of total costs) publishing net revenues should be more volatile (growing at a much faster rate than sales when sales are rising and falling at a much faster rate than sales when sales are falling).
Wiley revenues fell in 2009 and indeed you see that G&A increased as a percentage of sales in that year, but they continued to rise (as a percentage of sales) the following year as well – even though Wiley’s sales went up.
So what can we conclude? Well, at least in the case of Wiley, the internet has had little impact (yet ) on the volatility of Wiley’s net income. Therefore, fixed costs may be rising but the overall operational leverage of Wiley has not yet been significantly impacted by technological change. Frankly, I was so sure that Kent was right, I am a bit surprised by this result. The other remarkable result; I also thought that David’s comment about stability was spot on. However, the picture painted here, for this company, is one of great stability. Perhaps this is just a calm surface hiding dangerous rips underneath. One would have to dig into Wiley’s SEC fillings to find that out – which would require far more effort than I am willing to expend on this issue at this moment. But it was a fun little meander through the world of finance.
Looking at these numbers, I’m not surprised. COGS will fall because the variable costs are declining as print goes down. Fixed costs in staff are probably not captured in COGS but are in a salaries line that analyzing COGS or G&A won’t include. So the falling COGS fits with the notion that variable costs are falling as business migrates online, and the G&A numbers could be irrelevant if they don’t include salaries for all the functional areas. It’s hard to tell from these roll-ups.
Nevertheless, the concern isn’t too much for these big diversified businesses like Wiley, Elsevier, or WK, which have the negotiating power to outsource efficiently, the swath of revenue streams to blunt direct whacks at their expense lines, and so forth — my concern is for the small publishers, which suddenly have to deal with IT, new marketing, analytics, and management they didn’t have to deal with in the past. The shift of AIP out of the journals hosting business is throwing a harsh light on this for many of the smaller publishers affected. I’m hearing that they’re having to scramble to bring in-house staff to perform the services that AIP was previously providing. The fact that providing these became too expensive for AIP and now may be frighteningly expensive for the small publishers taking them back over is another indication how things have changed.
Yes, the figures you read in these statements are highly aggregated and hide a great deal. As Mark guessed, where publishers categorize IT expenses varies widely from one publisher to another. During my time at Penn State those expenses were subsumed in various departmental expense breakdowns, but then when we hired a full-time IT person, we switched to having an entire different section of our operating statement devoted to IT. Many smaller presses do not have any IT staff, but rely on the university’s IT people in other departments. Some presses, like ours at Penn State, benefited from their being an administrative unit of the library and having certain services, like digitization, provided at no expense to the press but counted as part of the library budget. The intra-university circumstances of presses vary tremendously, so it would be hard to generalize about any of this. You’d be better off getting access to data about stand-alone commercial academic publishing operations, like Lynne Rienner Publishers (for which i now work part-time, but have no access to their operating statement).
Mark, quoting from Wiley’s 2011 financials:
Operating and administrative expenses for fiscal year 2011 of $910.8 million were 4% higher than prior year, or 5% excluding the favorable impact of foreign exchange. The increase was primarily driven by higher technology costs ($15 million) reflecting ongoing investments in digital products and infrastructure, such as WileyPLUS, eBooks, customer data/relationship management initiatives, and Wiley Online Library; higher employment costs ($8 million) due to merit increases and retirement benefits; higher journal editorial costs ($7 million); higher warehouse rent and facility costs ($6 million) and travel expenses ($3 million) to support business expansion, partially offset by lower journal distribution costs due to off-shoring and outsourcing ($3 million).
Of those costs cited as driving the overall increase in expenses, most are fixed (technology, salaries, facilities), while savings occurred in variable costs (lower journal distribution costs).
Mark, a major firm growing 40+% in less than 5 years in an instability of its own, suggesting that the overall market dynamics are easily large enough to mask the impact of author pays. The first data I would like to see is just the number of articles published via author pays by year, plus as a fraction of the total published. I am not sure that the cost issues you folks are discussing are important in this context, especially if people start pouring money into author pays journals. We may see an investment bubble. In that case economics go out the window for the short term.
Kent, I agree that technology costs are more than likely not captured in COGS. In fact, working backwards from net income and COGS I got the very same $911 mil. for operating expenses reported in Wiley’s P&L. G&A would indeed include all salaries and marketing expenses, including those related to the new technology.
But the fact that costs are going up in absolute numbers is not the most important point. The most important point is looking at the numbers in relationship to total sales. We all agree that the industry is going through fundamental change. But the remarkable observation here is that in spite of all that change the relationship between costs and revenues has changed very little for Wiley during a very turbulent period. Kent, you rightly point out that many of the new costs are fixed. Yet, Wiley’s numbers do not yet reflect the increased volatility associated with increased fixed costs. Micro-economics is based on the simple concept of risk/reward. Industries with high fixed costs (mining, agriculture, airlines) have very volatile income streams; consequently investors generally demand a higher return from these industries than they would something like food and beverage. During this time period, for this particular company, net income is remarkable for its stability (both in absolute terms and as a percentage of sales). As you say Kent, it may be an issue of small versus large companies. If your theory is correct (and I think it is) you would see this issue manifest itself more visibly for smaller firms. It could also be that the time period observed here is also too short to reflect the new volatility in the marketplace.
David, you are right, author pay models and cost of publishing are two very different issues, the one is not related to the other. I don’t think it likely we will see an investment bubble in OA journals. Bubbles only begin when there are outsized rewards (usually caused by rapidly rising asset values). Asset values in publishing are not rising rapidly. Consequently, people aren’t going to pour money into publishing because there is more money to be made elsewhere.
Mark, I am thinking of a “dot.com” type bubble, not an asset based bubble. The former is fueled by rising potential return expectations. People are jumping into author pays, so it could bubble.
