As someone who has avoided buying Amazon’s e-books because of its restrictive use policies, it wasn’t a big surprise to see those policies eventually enforced in the form of termination of a customer’s account and deletion of her library for an unexplained violation of terms of service.

And like many others who have watched the divergence between people’s expectations of ownership of digital goods and the actual licensing practices around them, I’ve wondered whether anyone reads the end-user license agreements (EULAs) that define those relationships, much less understands them.

Given their length and complexity, common sense suggests no.   The notion that people might read them has spawned jokes and little tests of whether people are paying attention.  Ha ha.  None have been encouraging. The iTunes EULA is the butt of some of these jokes but is actually one of the shortest at under 2000 words.  (And no, at present, you don’t ‘own’ the music you bought.)  The Adobe Reader EULA is over 5000 words.  The Steam EULA is over 6000.  The Microsoft Office EULA is over 11,000.  Who would read all that?

Last year we had a chance to put this question to 2303 Americans and 1000 Germans.  We also asked whether they felt they understood the EULAs they read.  The results were pretty surprising.

That’s not a majority by any means–nor does it mean people read EULAs regularly or, for that matter, actually understand them–but it’s far higher than I expected.  I expected my own experience to be the overwhelming norm: I’ve clicked through hundreds of EULAs, glanced at a few, and carefully read… none of them.

So who are these EULA readers?  In the US, there are two main factors: the likelihood of reading a EULA increases with age, but the likelihood of installing software (and therefore encountering a EULA) drops with age.  80% of those under 30 have installed software.  Among those over 64, only 41% have.  So in terms of raw numbers, the middle-aged are the most frequent readers.  In Germany, the same trends apply but middle-aged Germans appear comparatively complacent about their EULAs–perhaps because the cases that have attracted attention in Germany have involved video games.

To date, US courts have been very sympathetic to the view that EULAs are binding contracts provided there is an actual moment of assent: a click through, rather than, say, just a link at the bottom of the page.  After that, anything goes.  And not just the overriding of various user protections embedded in copyright law.  Clicking through AT&T’s service contract, for example, relinquishes your right to participation in any class action suit against AT&T.  The US Supreme Court gave this profoundly anti-consumer behavior its blessing in 2011 (Scalia wrote the majority opinion).  Valve’s Steam license does the same.  With this ability to dictate the terms of redress, you can say goodbye to the use of the courts for corporate accountability. Will regulators pick up the slack?

German courts have acted more aggressively against EULA encroachment on user rights, ranging from requirements that the terms of any EULA be made available before purchase (a minor hurdle) to more substantive complaints against companies that fail to comply with Germany’s strict privacy laws.

EU courts appear to be following suit.  In August, the EU Court of Justice dealt a major blow to the core purpose of EULAs by reasserting the principle of ‘first sale’ for software.  Henceforth, software publishers in Europe will have to provide for consumer resale–a major change that is already creating friction with services like Steam.

None of this goes, yet, to the question of the validity of EULAs.  There are, of course, lots of contracts in daily life that contain fine print that nobody reads (probably).  But there should be limits on the range of rights that can be curtailed through means that, while formally accessible, are meant to be part of routine exchanges outside the context of negotiation and legal expertise.  This becomes more important in oligopolistic markets where customers have few choices if they don’t like the terms of service.  AT&T in the cell phone market, Valve in the game distribution market, ISPs, and many other vendors exercise considerable market power.  European courts seem to be going in the right direction on this  issue.  US courts are not.

We can connect this back to our usual concerns with piracy and digital media markets by viewing EULAs as part of the balance of power between vendors and consumers.  The more legal recourse consumers have against bad vendor behavior, the less likely they are to seek illegal recourse.  The Amazon attack is  important in this respect because it teaches new digital book consumers a simple lesson: not ‘respect our terms of service’ but rather ‘pirate your e-books.’ (Amazon’s “conditions of use” also forbid class action suits, btw).  If I were a publisher, this would be a strong signal to get out of the Amazon ecosystem and chart an independent path. As the music business has demonstrated, vendors don’t get many chances to set consumer expectations.

What does our data add to this conversation?  I don’t know.  I see no evidence that our  reported rates of EULA reading are a random outlier or the result of a badly formed question–unlike, for example, our question about movie file ownership, in which large numbers of people over 50 assumed we meant the files recorded on Tivos.  Possibly we asked the wrong questions–not ‘have you read’ or ‘have you understood’ a EULA, but ‘how much attention do you pay to EULAs.’

Gentle readers: this was a throwaway question expected to confirm preconceptions.  Instead it just raised more questions.  Help wanted.