The economics of paywalls

With on-line advertising revenues stagnant at best, and print in terminal decline, newspapers are starting to build paywalls. 

The economics of on-line media is a little different from the economics of print, and a lot different from the economics of, say, potatoes. The difference is shown in the diagram below. 

The diagram shows the amazing revenue potential of on-line distribution. A potato producer can only profit from his potato once. But a newspaper can sell an article a million times over. The question is: how can newspapers tap that potential?

Any business must solve two inter-related problems: how much to produce, and how much to charge for its product. For a potato producer in a competitive market, the problem is, at least in theory, straightforward. The producer charges the going market rate for his potatoes - that's the highest price he can charge and still sell his output. He produces as long as it's worth his while: the price he receives for his potatoes covers the marginal costs of production.

The on-line publisher's problem is a little different. First, she has to figure out how to generate the maximum possible surplus from production. Second, she has to work out how to get as much surplus as possible for herself.

The total surplus is maximized at point Q*, where the cost of producing one more article is just equal to the willingness of all potential readers to pay for that article.

Although publishers don't explicitly talk about marginal costs and willingness to pay, this is the calculation that is implicitly made. I know about this from managing the finances for academic journals. If the editor of Canadian Journal of Economics wants to publish an extra issue, the Canadian Economics Association has to weigh up the costs of each additional article published - copyediting, typesetting, mark-up, hosting - against the benefits. Will we get any new subscriptions? Will we keep more of our current subscribers?  Will people pay to read it? Typically the answer to those three questions is "no, no and no", but the Association is a non-profit organization, so weighs other benefits, for example, the opportunity for young academics to publish their research in the journal.

Once an on-line publisher has found approximately the right Q*, the question is: how does it raise revenue from consumers? Charging on a per-article basis is a non-starter. The only way a per-article charge can reach the surplus-maximizing point, Q*, is if consumers are charged their own individual share of the article's marginal cost. For example, suppose freelance, editing, posting and hosting costs amount to $500 per article on the margin, and the marginal article is read by 10,000 people. Then the per-person cost is $500/10,000 or 5 cents per article.

There are two reasons why such a charge is infeasible. First, it's so small that the costs of making the payment, from both the consumer and the firm side, dwarf the revenue the charge would raise. Second, if publishers are charging a marginal cost price, they will never cover their (quite substantial) fixed costs of, say, software, hosting, and so on. Yet a per article charge of, say $1 per article would reduce the quantity demanded - and thus the consumer surplus - far below that available at Q*. The benefits of the entire publishing enterprise would evaporate.

The alternative is to erect a paywall, and charge consumers a fixed monthly fee for unlimited access to the site's content. The maximum amount a consumer would be willing to pay for a subscription is equal to the entire area underneath her demand curve - that's the benefit she receives from subscribing to the on-line service. 

In some sense all of this theorizing just re-states the obvious: publishers should produce a good product, and charge the maximum the market will bear for it. Yet it can be used to shed light on real-world publishing dilemmas.

For example, PostMedia has erected paywalls around all of its local papers - the Vancouver Sun, Calgary Herald, Regina Leader, and so on. These papers are substitutes for one another, carrying the same national and international news stories. A person's willingness to pay for each individual newspaper is less than their willingness to pay for all PostMedia newspapers - why should I pay to read a story in the Citizen if I can get it as one of my 10 free monthly articles in the Vancouver Sun? Hence if PostMedia wants to maximize profits, it should institute a PostMedia wide paywall, rather than paper-by-paper paywalls.

The analysis also shows that there is a direct trade-off between people's willingness to pay for subscriptions, and the number of free article allowed each month. If people are allowed, say, 20 free articles each month, the willingness to pay for a monthly subscription decreases by the entire area to the left of the point marked "20" in the diagram above. The amount that can be charged for subscriptions is people's willingness to pay for the articles they could not otherwise obtain for free.

So far I have been discussing pricing strategies - the best way to charge for content - assuming that the entire publishing enterprise is viable. But what if it isn't? What if that yellow "total willingness to pay" area is less than a publishers' fixed costs of doing business?

The internet erases space. A publisher has to convince a potential subscriber that his or her paper provides more valuable content than the next best alternative anywhere else in the world. Thirty years ago the Ottawa Citizen competed with the Ottawa Sun and local television for eyeballs. Now it competes with the New York Times, the Economist, Slate, the Atlantic, the Financial Post, as well as bloggers, tweeters, and so on. For a publisher who captures a global audience, the internet is a golden opportunity, but for others?

Nothing succeeds like success. A newspaper with a strong base can share the cost of creating content across millions of subscribers, giving it huge advantage over its rivals - revenue generates content generates more revenue. 

But what about local newspapers? Yes, new technology and the internet allow for some cost savings - managing and copy editors can be dispensed with, for example - but only so many. One view is, if newspapers can't make profits, then they had better fold, and allow resources to be deployed more efficiently.

However alert readers may have spotted that the analysis of newspapers looks a lot like the standard economic analysis of a public good. Even when people value a good, like local news, they have an incentive to free ride, to let others pay the cost of collecting and publishing information. The internet sometimes seems thousands of miles wide and just a few inches deep - all of the blogs, local web pages, and so on rely on the hard news gathered by serious journalists. The marginal cost of opinions is pretty low - everyone has one. The marginal cost of reading through omnibus bills, filing freedom of information requests, making cold phone calls, tracking people down and asking difficult questions is a lot higher. If the fourth estate fails, who will take its place?

Here are some questions on the economics of paywalls. These would be suitable for an undergraduate economics class or graduate public policy class studying public goods: Download Economics_of_paywalls_questions

Comments

[JR Hulls]

A very thought-provoking post, but one can not claim that commercial publishing is primarily in support of a public good, which is almost a secondary benefit. Cynically, the news, and other coverage is there so people will read advertisements.

There is a very interesting parallel case with academic publishing and the revolt against the rising cost of access to scientific papers. This is epitomized by the recent announcement by the Wellcome Trust to penalize scientists by withholding final grant payments if their papers are not open access. http://www.guardian.co.uk/science/2012/jun/28/wellcome-trust-scientists-open-access

Here, the primary driver is the value to the researcher of access to literature in his field. An individual, not associated with an institution can expect to pay a substantial sum for temporary access to a needed paper, and several universities have revolted against the publish for profit motive of the major journals, as set forth in Winston Hyde’s statement as to why he resigned as editor of the prestigious Genomics journal. http://www.guardian.co.uk/science/blog/2012/may/16/system-profit-access-research

I would be interested in your analysis as to how the actual content value of news drives the economics, as well as the impact of the incredible noise/signal ratio of the internet, where it seems we’re heading in the direction posited by Negroponte in his classic “Being Digital” as discussed here: “Reading the Daily Me at the End of the Century of Self” http://somewhatlogically.com/?p=311

[Frances Woolley]

JR "one can not claim that commercial publishing is primarily in support of a public good"

Perhaps I should have been more clear that I was using the word public good in the economists sense, that is, a good that is non-rival and non-excludable. On-line publications are non-rival but not - to the extent that one can put up a successful paywall - nonexcludable.

Scientific journals are an interesting example of the way in which the economics of paywalls can be used to generate substantial profits. The subscriptions are paid, however, by university libraries, especially those in the US, and they're starting to play hardball - geting organized into state- or regional-level buying coalitions and flexing their monopsony power. Or saying "sorry, can't afford it." For this reason, and because of pressure from funders of the type you mentioned, I think we may start seeing changes in this market too. I don't know.

I have no idea why people demand what they do, so no brilliant insights into the link between content and econoics.

[david]

Content piracy is a problem; for any two individuals that would pay $5 each, a devious third party could charge them $4 each instead and pay $5 to the original producer, earning $3 profit by driving the original producer's marginal revenue ever closer to marginal cost (eventually, zero).

This dynamic becomes particularly obvious and sadly observable near the bottom, where 'costs' are imposed by obliging users to watch advertisements.

The handful of successful content sellers have had to rely on technologically-enforced bundling with services rather than trying to sell content alone, really. It's just too easy for the anonymous public to duplicate your content.

[Donald A. Coffin]

I might note that one implication of newspapers moving to on-line publication is likely to be a *return* to paying for content, rather than the "newspaper" being primarily a vheicle for advertising. Befoe about 1850 (I'm guessing a little there about the date), newspapers were largely content, with little or no advertising. As local, and then ohter, businesses recognized the potential of newspapers as a means for communicating with customers and potential customers (or as newspapers realized that potential and reoriented their business plans), direct reader revenue fell as a percentage of total revenue. One implication of this analysis is that this will reverse.

[Frances Woolley]

David: "The handful of successful content sellers have had to rely on technologically-enforced bundling with services"

I'm not sure exactly what you mean. My motto when managing the Canadian Economics Association's finances was "membership has its privileges". People's willingness to pay for the Journal was close to zero (as most could download the papers from university libraries, or access them in working paper form) so we kept membership levels up by having differential conference registration fees for members and non-members. Bundling with services in other words. But I can't think of a lot of similar examples.

Donald - interesting.

[david]

Bundling with social networking features, for instance, so that commenters use their subscription account to post. So to post where all the other commenters are posting, you have to have an account that the site can control.

It is probably for this reason that many prominent paywalled sites don't even bother to rigorously protect their original content alone to begin with: browsing NYTimes.com or Economist.com indefinitely is as simple as clicking a bookmarklet to clear cookies. Why bother throwing up more than a cursory barrier, which a content pirate could easily surmount? People are actually paying for other reasons besides pure access to the content.

[oblivious]

The Guardian has an affiliated dating site they sell subscription to in a bundle. If you subscribe to The Guardian, maybe reading dating another Guardian subscriber is more likely to work out for you.

Profit isn't the only utility of owning a news site (or generic article aggregation site). You may like influencing opinion (partisans or unions), you may be in it for the prestige (Conrad Black?), or you may do it to promote something you believe in (advocacy groups, academic publications). In this context many publications can survive simply by keeping costs low and trying not to lose too much money.

My prediction is that the naive paywall model can only be sustained by a relatively small number of highly prestigious publications at any time: there's too much competition from people willing to lose money and entrenched players who can undercut you while making way more profit.

[Frances Woolley]

Interesting examples of tie-ins, though interestingly the Guardian doesn't charge for news, only for soul mates (I guess - I didn't get to the point where I'd have to pay).

Oblivious - I guess this gets back to the question of the determinants of demand - are people willing to pay for news when people with an agenda are prepared to offer slanted infotainment for free?

david - Actually I think lots of people are pretty lousy at getting around paywalls, and even if they can they don't - it feels a bit like cheating.

[Thomas]

Frances,

I think related to David's point is that for many types of publishing an effective paywall makes the product less valuable to subscribers -- it's the opposite of a rival good in that sense.

That's familiar in the academic setting: publishing in a journal few people read is less useful, and so being able to reading articles in a journal few people read is also less valuable. Now we have blogs, the same sort of thing is true for newspapers. Felix Salmon has contrasted the Financial Times and Wall Street Journal paywalls, which make for unfriendly linking, with the NY Times paywall, which encourages links -clicking on a link to an NYT article isn't 'getting around the paywall'. He subscribes to all three of these, but it's harder for him to direct people to FT and WSJ articles, which reduces the value of those subscriptions. In yet another field of publishing, books that you can't lend to your friends are less valuable.

I'm not sure exactly what the anti-rival nature of publishing means to the analysis, but the NY Times paywall intuitively looks like a good response: subscribers can read as much as they like, and they can show anything they like to non-subscribers -- which is not that dissimilar to the way newspapers worked in the days when we had photocopiers but no web. The subscription cost still needs to be related to the fixed and marginal costs in the same way, but the marginal value of an article can't be treated as entirely independent of the cost and accessibility.

[Incidentally, I don't know what has caused the recent increase in Frances-blogging, but I'm definitely in favour.]

[david]

Prohibitively high technical incompetency and high weights on non-cheating is inconsistent with very high estimated rates of pirated software use...

[Bob]

It sounds to me like an argument for funding the Public Broadcasting System.

[Frances Woolley]

Thomas - interesting remarks on the ease of linking. I've started using twitter more and facebook less because I'm finding facebook's linking apps more intrusive. I don't know if "rival" is as good as word here as substitute, or if you're getting at is actually a form of network externality - people want to subscribe to the paper that everyone's talking about. Thanks for the comments about the increase in blogging - the MB curve shifted up on this one because my students have an assignment on paywalls.

david - it's a big world out there, lots of people pirate, and lots of people don't. There's also a difference between pirating a high-cost product from a corporation one doesn't like very much (Word) and pirating a low cost product from an organization one would like to support.

Bob - I think we're going to start to see, in the next little while, whether or not there really is a serious market failure needing government provision or finance. The cost per reader of delivering content is generally going to be higher in the Canada than the US, because we're smaller, which leads people to read US publications because they're better and cheaper, which means....

[Jacques René Giguère]

I sometimes buy the G&M for one article.The rest is essentially free to me but of negligible value. A paywall of $ 10 a month is higher than my monthly benefit. I don't buy the digital subscription. G&M loses $ 7. What do they gain?

[Determinant]

Newspapers and other print media are unusual in that they have two simulteneous customers: advertisers and readers. Their business model involves a mix of having the reader pay some portion (the subscription price or the $1.75 for the day's paper) plus advertising revenue, which consists of selling access to their readers. That's why actual news content is referred to as "the news hole" in the trade, it's non or low-revenue space.

Maclean's or the Globe and Mail raise some nominal fee from subscription revenue, but that hardly pays for John Ibbotson or Jeffery Simpson. The real money comes from advertisers. Paywalls are just an example of shifting the burden more to the reader. It will be interesting to see how this works out because when you purchase TV shows onine (legitimately) you purchase them without commercials, that's part of the deal. Will readers want to have advertising in a space they have already paid a fee for?

[Bob Sherman]

Newspapers once seemed to be vehicles for the delivery of ads. The news stories and their slant were what made one newspaper preferred over another. The profits were not made by subscription costs (which probably covered the cost of the paper and the devivery system), but by ads.

Now the delivery system is electronic, virtually free. Why is it so hard to carry on making profits from ads--be they store coupons, items for sale, or personal dating ads?

Shouldn't the objective of online media be to attract more readers with free content rather than to limit readership to paid subscribers? The more eyeballs you capture the more you can charge for advertising.

[Jacques René Giguère]

Bob Sherman: because Craigslist or Kijiji can deliver the ads without bothering with the news. What I discovered by teaching is how few people really want to learn...

[Chris J]

Frances,
I have filled bloggers' tip jars on occasion and was a member of NPR in Pasadena. We will start paying for things. For news, I think we will have a market failure with public broadcasters stepping in in some places and a membership-based system in others. I can see the day when the western world gets its news from the BBC, NPR, Al Jazeera, and a handful of smaller outlets (CBC here, ABC in Austrialia). Local news will come from small local papers like Northern Life here in Sudbury. In short I see news getting both more global and more local.
Why on earth would I pay a paper for economic analysis when I can read WCI? As Delong would say "Anyone? Anyone? Bueller?"

[Frances Woolley]

Determinant: "The real money comes from advertisers" - yup, and it's not clear how well on-line advertising (with the exception of paying for high search results) works. Perhaps people's willingness to pay for news just isn't enough to pay for the current news model - that seems to be the argument Tim Worstall makes here http://www.forbes.com/sites/timworstall/2012/09/24/why-i-think-the-new-york-times-is-going-to-go-bust/ (thanks for the shout-out Tim!)

Jacques Rene's comments seem to support the people-just-aren't-willing-to-pay-that-much theory.

Chris J - I wonder if we should put a tip jar on WCI. The question is what we'd spend the money on... I think it would have to be a booze-up somewhere

[Frances Woolley]

Chris J - I agree with you about the more local *and* more global comment.

[Jacques René Giguère]

Booze-up? I suggest The Mill,corner of Wellington and Portage Bridge. The poutine with St-Albert cheese is one of the best I ever tasted...

[Determinant]

Well, the current news model is breaking. It used to be that local TV stations were a steady business and good to own, now they are going out of business. Media as a money game is rapidly fading, and media as a power game is fading too, much to politicians' and media owners' regret. I know young political operatives whose job is to clip media headlines about their bosses. I don't think their bosses are too alert to the Internet, but they should be.

The whole edifice of the CBC/CTV/Big newspapers is fading in the face of downloadable content and the internet. If you think Can-con regulations create a deadweight loss, then it's an economist's dream. CBC already offers plenty of TV shows free for download, they have some ads but not on the order of TV channels.

Between the consumer-friendliness of downloadable content and the costs of traditional broadcasting, I think we'll see an end to traditional broadcasting in my lifetime. The CRTC as King of Media is also likely done as a model. Sorry Konrad Finkelstein.