Dan Lewis

Rupert Murdoch’s interest in erecting paywalls around NewsCorp content and removing it from Google’s index is in the news seemingly daily.  And today, it looks like Google made a play to keep the Wall Street Journal and other NewsCorp properties in the index.  Mashable reports on the two changes, and the second one has interesting fair use implications:

Publishers now have the option to tell Google’s spiders to only crawl and index the “preview pages.” This refers to pages that display the first few paragraphs of an article on subscription sites like WSJ.com in order to entice them to pay for a subscription. If a publisher chooses to have spiders crawl their articles in this manner, they will be labeled with “subscription” within Google News.

If you don’t know what fair use is, there are four factors that courts look at to see if an otherwise infringing use of one’s copyright is “fair” and therefore non-infringing:

1. the purpose and character of your use
2. the nature of the copyrighted work
3. the amount and substantiality of the portion taken, and
4. the effect of the use upon the potential market.

For more, this Stanford guide is a good start — indeed, the bullets above are a direct copy from there — but if you can handle some legalese, you are best off reading a case.  I recommend Warner Bros. Entertainment Inc. v. RDR Books (“RDR“, for short), for two reasons: (1) it’s about Harry Potter and (2) it shows that fair use is not at all intuitive.  The case doesn’t apply here but it’s easier to slough through than most cases.

Anyway: The fair use analysis.  I am going to skip the second factor because I think it’s basically irrelevant in this case.

Factor 1:  Is Google’s use transformative?

Yes, generally.  NewsCorp provides the content to report the news.  Google uses the content to provide a research tool.

While Google would almost certainly agree to not index NewsCorp content irrespective of the legal question, I wonder if NewsCorp could force the issue legally.  The answer: Probably not.  Crawling and indexing is transfomative: the search service (Google, Google News, etc.) is a research tool.  This article about litigation over Google’s Image Search explains more, and the analogy should hold.  In that case, Google argued that it provides the image thumbnails in order to allow it to act as an effective research tool.  The court noted that Google isn’t providing the full-size image, but rather, just enough for the searcher to see if they’re on the right track and click through.

If Google provided access to the full text instead of the synopsis, it seems that they’d be hard-pressed to claim they are providing a research tool  This factor would be in play and likely tilt toward NewsCorp.

Factor 3: Is the amount and substantiality of the content “too much”?

Let’s look at academic papers and their research tools; specifically, PubMed.  Check out this search result for “food protein induced enterocolitis syndrome” and you’ll see it’s a mere syllabus.  It’s substantial enough to meet the needs of the researcher, and the amount — well, the copyright holder wrote the passage so that researches can find the paper.   It’d be hard for the author to argue that using this is “too much,” but it’d also be hard for the research tool provider to argue that this is not enough.

The parallel to preview pages, written by the publisher (NewsCorp, for example) for the benefit of the researcher, is stark.  If NewsCorp provides these pages to Google (and the pages are indeed adequate to meet the researchers’ needs), Google would be very hard pressed to make a fair use argument for indexing the content behind the paywall.

Factor 4: What is the effect of Google’s use on the potential market for NewsCorp’s content?

Seems like an easy one. If Google indexes the preview page, the effect is minimal at worst. If Google indexes the article itself (and provides it for free), different story.  NewsCorp is drawing a line, albeit maybe one in the sand, saying that there’s a paid-for market for their complete content.  If Google crosses that line, the factor tilts into NewsCorp’s factor, quickly.

So, I think that at the end of the day, Google is doing what it has to do, not what it necessarily wants to.

The discussion around my post yesterday also entailed this comment over on another blog, and got me thinking further about what one actually purchases when one buys and e-book.  The same, of course, applies to mp3s and any other property which can be reproduced by a third party at very low (often no) marginal costs.

There is a disconnect of language here, probably a side effect of legacy businesses working with their legal teams to try and grab control while the consumer base, disorganized as it naturally is, is expected if not forced to make its arguments with the rubric set by the producers.   In other words,  “buying” an e-book is different than “buying” a book, even though from the consumer’s standpoint, it shouldn’t be.

Let’s start with tangible books — you know, the dead-tree versions that sit on shelves.  For those, there’s a pretty clear bundle of rights and lexicon, which I articulated in the comment linked-to above:

If I buy (“own”) a book, I expect to be able to do things such as re-sell, loan, rent, gift it. If I rent or borrow (“posses”) a book, I don’t, but expect to be able to do things like take it with me on a trip. If I am in your house and flip through (“access”) a book, you being a mensch aside, I probably can’t just walk out the door with it.

That’s not complete, of course, but it’s intuitive.  For centuries, we have culturally understood ownership to mean something absolute, constituting exclusivity and control.  If I own it, you don’t.  You can’t tell me what to do with it, what not to do with it, etc.  Obviously, there are going to be some limitations on my ownership when my ownership rights conflict with something in your bundle of rights in something.  But the meaning of the term is pretty clear.

What is also typically clear is how one gets to own something.  In most cases, you either buy it or receive it as a gift/inheritance.  Sure, there are other situations, but even then the transfer of ownership is most often clear, clean, simple.  Just over five years ago, two friends were over my apartment.  As one left, he noticed a DVD on my bookcase, and asked if he could borrow it.  I of course said yes, and he left.  My other friend commented, immediately after, that I would never be getting the DVD back.   We both knew he was right (and he still is).  What lawyers may deem theft by conversion I instead saw as acceptable if not annoying.  But in any event, it was clear — by any definition other than a hard legal one, my friend now owned that DVD.

Indeed, the biggest virtue of “ownership” is that it’s simple.  Everyone — even a toddler (“Mine!”) knows what it means.

E-books, specifically, and digital media, generally, muddle that up.  Right now, when you buy an e-book on your Kindle, you most definitely do not own that “book” in the typical meaning of the word “own”.  You cannot, lawfully and/or technologically, use it in the way(s) which you would have been able to use the paperback version of the same content: you can’t lend it to a friend, donate it to a library, re-sell it, etc.  Your rights are clearly delineated, I’ll bet, in the licensing agreement you entered into, but as the consumer, that’s not the bargain you expected. What is expected is the simple, common language: I bought it, therefore I own it.

The troubling part about all this is that the time-tested concept of ownership is clear, yet we have some odd expectation that it will yield to granularity in licensing.   You can buy DRM music in the iTunes store or non-DRM.  Amazon uses DRM-free content as a sales point.  On the other side of the aisle, Creative Commons has six different licenses, some of which use ther term “non-commercial.”  What does non-commercial mean?  Answering that required a year long study (in which I participated) to yield a 255 page (!) .pdf to “define non-commercial”.

It’s all way, way too complicated.  We need to find a way to keep it simple.

Imagine this: You could walk into a building and borrow, rent, or buy used books.  We could call these buildings “libraries” or “used book stores”.   Heck, while we’re at it, let’s also imagine people giving books that they’ve just finished to their friends.

If you look at the copyright law — 17 U.S.C. 106 — you may be surprised to see that the law explicitly says that these transactions cannot occur, at least without permission of the copyright holder.  Really:

[T]he owner of copyright under this title has the exclusive rights to do and to authorize any of the following:

[. . .]

(3) to distribute copies or phonorecords of the copyrighted work to the public by sale or other transfer of ownership, or by rental, lease, or lending.

See? Illegal, all of it.

But of course, we know better.  We know that libraries and used book stores exist, lawfully so at that, and (correctly) discount the idea that these institutions have cleared copyright with every single author and publisher whose work appears in their inventory.  We also know that no one gets in trouble for passing a book to a friend.

That’s because of the First Sale Doctrine.  Originally stemming from a 1908 Supreme Court case and codified at 17 USC 109 in 1976, the First Sale Doctrine, in a nutshell, says that the copyright holder only holds rights of distribution in order to put the product into the stream of commerce.  Once the product is lawfully sold once, the right of distribution evaporates.  So yes, go ahead and send me copies of books you think I should read.

But what about e-books?

Imagine, then, this:  You could fire up your laptop or Amazon Kindle and borrow, rent, or buy a used book.  Heck, while we’re at it, a friend could lend of give you a used copy of their book, too.

Right now?  Unlawful.  An exception to the First Sale Doctrine may render it irrelevant, and in any event, the digital transaction necessarily involves copying — thereby impinging on the copyright holder’s exclusive right at 17 USC 106 (1): “to reproduce the copyrighted work.”

Seems silly, but that’s the lay of the land.

Both John Battelle and Mike Shatzkin have posts today about e-books, and both touch on this.  Shatzkin discusses best practices for e-book publishers to combat piracy while Battelle complains about features his Kindle lacks:

You can’t share a Kindle book with anyone else. That’s just nuts. The sharing of a book is perhaps one of the most intimate and important intellectual acts between humans, ever. I’m not stuck on whether or not that sharing is physical. I’m stuck on the inability to share. It’s a crime.

The juxtaposition shows the problem.  The consumer (Battelle) demands the ability to share (“It’s a crime” that he can’t!) while the publisher (through the eyes of Shatzkin) is desperately trying to maintain control over the   distribution of e-books, with the effect of controlling downstream distribution as well.   These goals are irreconcilable, so something has to give.

Which brings me to the used book store in the cloud.   Say an annual fee — $50? $100? — gave you the ability to rent a book directly to your Kindle or other device.  The fee could be built into the price of a Kindle; as well as selling the current version for $259 without this service bundled, Amazon could offer it for $200 plus a $100 annual fee and multi-year contract, much like a cellular phone.  Either way, one would still be able to purchase — permanently — a copy of an e-book whenever they see fit.

The books would actually live “in the cloud” — that is, on the publishers’ or distributors’ (Amazon’s? B&N’s?), and be pushed to your device as-needed, with some liberal allowances made for things like flying, etc., when wifi and cellular connectivity is lacking.  Copying the file could be made difficult, especially if the service were limited to e-readers such as the Kindle.  You could have a certain mumber of rented books on your device at a time, say three, but maybe fewer.   And while your friend could not just give you a book he or she purchased, they could definitely recommend them to you via some sort of social network-like interface.  But each of these digital copies would exist in a finite space, as determined by the publisher and policed by the distributor.  Quite literally, we’d be applying the First Sale Doctrine to e-books as much as possible by re-introducing scarcity to (e-)books.

For Battelle, this would allow him, effectively, to share books with friends, so long as his friend was part of the renting ecosystem.  As the annual fee would be easily incorporated into the Kindle’s price, this would be even more palatable.  Further, I bet that one of his other gripes — that you can’t put the actual, physical book on your shelf — would be mitigated as the marginal cost to acquire the book would be offset in kind.

For Shatzkin and publishers, this would massively reduce the incentive to “pirate” e-books, especially among those consumers who would be likely to purchase books if otherwise available for a rental/browse/borrow discount.  It would certainly cut into the number of sales of e-books, but because of the lack of permancy, not into the sale of paper-and-ink books.   It probably is a net negative compared to the revenue stream provided by a world without e-books at all, but that universe is gone and not coming back.

The irony?  While the publishing world looks toward saving the present structure, they’re doing so by ignoring the First Sale Doctrine — a common-sense point of law in both contexts.  Adopting a First Sale Dotrine for e-books would be a huge step toward meeting everyone’s needs.

The Office is a brilliant TV show, but you probably already knew that.

But did you know that there’s a three-part “Webisode” on Hulu? If not, take ten minutes — actually less, including the ads — and watch all three, below.

Episode 1: Creative Differences (2:10)

Episode 2: The Replacement (2:13)

Episode 3: The Music Video

A few questions:

1. Does it cost more to make seven of these or one full episode?  Ignore distribution costs — I care more about production, including cast salaries.
2. How much does it cost to produce one of these episodes relative to the cost of producing and distributing a television ad?
3. Compared to a full episode, which is more likely to build an online following?
4. Compared to a TV ad, which is more engaging?  More “viral”?   More likely to be sought rather than skipped?

If the answers to the above skew in the direction the title of this post implies, there is probably a huge opportunity here.  Imagine, if you will, if Terry Tate: Office Linebacker debuted today.  In 2003, before Hulu and YouTube were here to make online video distribution easy, and before Facebook and Twitter were around to make content spread from friend-to-friend, the flagship video witnessed 7 million views on Reebok’s website.   Last night’s episode of The Office, by comparison, drew roughly 8.15 million viewers.

Yes, the numbers reek of an apples-to-oranges fallacy; and yes, creating a Terry Tate is hard.   You’d have to hire writers.  You’d have to develop characters.  Actors would not just stand there and smile, but deliver lines and interact with each other.  Permanent sets would need to be built.  The list, I’m sure, goes on, and again, it’s not easy.

But if you can do it, it’s invaluable.   Imagine how amazing it would be if consumers not only watched your ad, but made a point of watching it — every single day.  And then shared it with a friend.

Two of the shows I watch on Hulu — House and It’s Always Sunny in Philadelphia — hit the website with an 8 day lag from their first television airing. That is, when an episode of House airs on Monday night on FOX, it won’t be available on Hulu until the day after the next episode airs a week later.  The same is true for Sunny, which airs on FOX-owned FX, but on Thursdays.

I understand why FOX does this.  If you are a true fan of either show, you likely won’t want to be a week behind, and will only use Hulu has a backup (if that; DVRs are better), and the television medium monetizes better than web/Hulu.   And the potential audience for a TV show is huge — just look at well, House itself.  It’s on a major network with a huge marketing budget. It’s Facebook page claims almost 2.5 million (!) fans. It does really well in the TV ratings game and pulled 16.5 million viewers for this season’s premier.

But the eight day lag, especially in Sunny‘s case, I find two days too long.  Why?  Because I’d gladly watch Sunn

on TV — if I could only remember to.   And I remember to do so every week — on Friday, when last week’s episode hits Hulu.  So I watch that episode, laugh hysterically, and then consider DVRing next week’s show.   (NB that the FOX-demanded Hulu time shift requires that I, too, time-shift my TV-based consumption; that is, it’s DVR or nothing as far as TV is concerned.) Then I remember something else — I’m a week behind.  So, if I DVR next week’s show, I’ll have to watch this week’s next Friday on my computer like I usually do, and … bleh.  So I stay wed to Hulu.

The fix?  Easy: lag the shows by six or even seven days.  On Thursday night at 8 PM, I watch Sunny on Hulu and am caught up.  Two hours later, I’m watching on TV.  I log into Hulu next week and see the show there — a re-run, to me — and remember, hey, it’s on again tonight.  Use Hulu as an advertisement, and you win. Big.

Marc has a well thought-out takedown of a pair of Ethicist posts by New York Times columnist Randy Cohen on ticket scalping.    Cohen defense Miley Cyrus’ practice of instituting anti-scalpling measures by asserting , basically, that not everything should be auctioned off to the highest bidder, and performers should be allowed to license out the right to listen to their performances.

But why would they?  Artists can easily ride the demand curve, auctioning off tickets or otherwise maximizing revenue.  Even if they do not have the ability to do it themselves, they can partner easily.  Take Major League Baseball: They have an agreement with StubHub, allowing the latter to re-sell tickets, with the league and its member clubs taking a cut.  Why would Cyrus not do the same?  Is she really thinking of her fans?  If so, she’s spending a lot of money — in lost revenue — to be “fan-friendly.”

No so, Marc argues.  Rather, and correctly, I think, he proposes that artists are more interested in selling out venues than maximizing revenues.   Why?  Because failing to sell out a show signals a lack of popularity, and in their world, popularity matters.

Which is reinforced by another thing that can happen on the secondary ticket market: Ticket prices can actually fall below face.  I spoke to Mark — yes, another Mark — an experienced ticket arbitrager, and found that prices actually fall more often fall below face value more often than you think.   He says that Cyrus, who is popular everywhere, is an extreme outlier: even tickets to Britney Spears’ shows in sparsely populated areas during her recent tour could be obtained at face (or perhaps under) on the secondary market.  And if that were to happen, again, we’d have a signal of waning popularity.

Cyrus is participating in the latter market, which is a market for an even scarcer commodity: attention.  Popularity fuels attention.  Attention leads to listeners, viewers, and over time, more and more revenue.  Losing attention is too risky, and scalpers put it at risk.

[Update: June 5, 2010: In January, an armed robber held up a jewlery store on Madison Avenue between 75th an 76th, shooting and killing a 71-year old employee. This is the first shooting homicide in the Upper East Side in over seven years, and per the linked-to article, seemingly random.]

The New York Times came out with a neat interactive map of homicides in New York City — something which would be really cool if it were not about murder.  According to the map, ove the six and a half year period covered (2003 to present), NYC recorded just over 3,400 homicides, of which 15% — or about 500 — were in Manhattan.

The Upper East Side section of  Manhattan is incredibly safe, apparently.   11 homicides over the period, or about two per year.  For an area with a population of 200,000+, that gives roughly a 1 homicide per 100,000 per annum.  To put that in perspective, in 2003, Maine had the “best” homicide rate in the nation, at 1.2 per 100,000.

There are no formal barriers separating the Upper East Side from the rest of Manhattan — yes, it abutts Central Park on the west but one can easily walk though that, and the east is bound by the East River.  North and south, though, the lines are 96th and 59th streets respectively, which is to say, the borders are entirely fluid with the neighborhoods next to them.   (A guess as to why the lines are where they are?  On the south side, Central Park ends on 59th; on the north, the Metro North train tracks come above ground at 97th.)  The borders are known but hardly observed in any true fashion — we regularly consider a playground on 99th and 5th to be in the “Upper East Side”, for example.

If you live here, you probably will understand why the neighborhood is safer than most, but experience does not translate well to words.  I was not terribly surprised to see a low homicide rate, although admittedly it was much lower than I expected.  What did surprise me, however, was one simple fact: no firearms were used in the homicides.  Or, more accurately, guns abide by the borders.

The map above sums it up.  The green dots — firearm-related homicides — all fall outside the borders of the Upper East Side.  There are two which are close, of course: 57th and Madison, 97th and 3rd.  Two blocks south of the UES, one block north.   This makes the UES a massive outlier, as 69% of the 3,400+ homicides city-wide were firearm-related.

Why?