A response to Steve: It still doesn’t sit well

by on May 9, 2013 · 3 comments

Over at Freedom to Tinker, Steve Schultze has a response to my Reason article about Craigslist suing its competitors. Steve expresses some surprise that I would suggest that we might want to recognize a new property right since I have been so critical of the excesses of our current IP regime. Let me take a stab at reconciling that seeming paradox.

First, I should say I’m sympathetic to Steve’s position, which he shares with many others, and which may well be right. I wrote the Reason article more than anything to provide some balance to what I saw as a knee-jerk reaction in the blogosphere to the Craigslist ruling. I really didn’t see anyone giving Craigslist’s claims a fair shake (probably because the company is acting hypocritically given the public profile they have cultivated). That’s why in the article I’m ambivalent about whether Craigslist should have any remedy, and why I don’t make the case that trespass to chattels is the right approach. The point is that neither am I convinced that it’s clearly the wrong approach, or that Craigslist should clearly not be waging this suit.

That said, let me suggest that my thinking on this is not at odds with my thinking on copyright. Steve chides me for saying that maybe there’s something to Craigslist’s claims because what its competitors are doing doesn’t “sit well.” He says that “the notion that something doesn’t ‘sit well’ is not necessarily a good indicator that one can or should prevail in legal action,” and he’s right, which is why I don’t make that claim in the article. He goes on to admit that “to be sure, tort law (and common law more generally) develops in part out of our collective notion of what does or doesn’t seem right.” And that was my point. The fact that what Craigslist’s competitors are doing doesn’t sit well, I suggest, should give us a hint that this isn’t as open-and-shut a case as some have made it out to be, and that perhaps we should take a closer look.

I’m glad Steve brings up the common law. One of the central critiques I have made about copyright as a property right is that it did not develop at common law, and is instead a creature of statute. The fact that copyright is created by politicians guessing about the future (and influenced by special interests), rather than courts deciding actual cases and controversies, is what in large part leads to its excesses. I am much less skeptical of property rights that emerge at common law over time after an evolutionary process of trial and error, and as Steve points out, this process usually begins when a court is presented with a novel question that doesn’t “sit well.”

On to nitpicks about copyright. Steve says that “there is a bit of confusion around the copyright claims” that he wants to clear up and notes that, “The court held that Craigslist unambiguously does not hold copyright in user-created postings, except for those three ill-fated weeks last summer when they instituted that horrible terms of service.” That is a confusing statement in itself because at face value it implies that Craigslist does indeed have copyright over the postings created during the couple of weeks it required users to click through what Craigslist claims is a notice that assigns to it an exclusive license. Of course, as Steve knows, the court’s ruling was on a pretrial motion and whether Craigslist has copyright in that small set of listings has not yet been determined. And if I had to bet, I’d bet the click-thru won’t qualify as an assignment of an exclusive license.

He goes on to say that “Jerry’s claim … that no copyright exists in these posts whatsoever, seems weak.” Well, again, I made no such claim. The only thing I said on this was that “a site like PadMapper only copies facts about a listing (i.e. 3 bedrooms, 800 sq. ft., $2,000 a month, etc.), and mere facts are not subject to copyright.” I wasn’t suggesting that the listings in questions were Feist-style facts that could not be copyrighted, but that PadMapper is not copying the listings wholesale, but only taking mere facts, which would be unprotected given the merger doctrine. Steve seems to miss that distinction.

Even if I had said that the listings were not subject to copyright, Steve’s evidence that such a claim is weak is, well, weak. He points to the court’s ruling, which finds that the listings probably are copyrightable. Yet as Steve notes, the facts in the case are read in the light most favorable to the non-moving party, which is Craigslist since this is a motion to dismiss. So I’m not sure how a court looking at the issue in a light favorable to Craigslist and finding that Craigslist has adequately alleged sufficient facts to proceed is much evidence against the legal claim that the listings aren’t copyrightable.

Anyhow, after clarifying “confusions around the copyright claims” that I did not introduce, Steve takes me to task for “argu[ing] that maybe this is the right case to test out some novel approaches to applying physical-world torts to online things that feel kinda like property.” In this he’s probably right; this is likely the wrong case. It is “a messy fact pattern,” as Steve says. I think that’s true in part because Craigslist has not been adding new features to its site and has seemingly been resting on its laurels since the late 90s. A better case would be one that involved, say, a service that took Match.com and OKCupid listings without permission. It would be a much more clear-cut question if we had a plaintiff that was working hard to attract users’ listings only to see them taken without permission, as 3taps says in its white papers should be allowed. Alas, though, this is the case we have.

Steve then goes on to the heart of the matter: my suggestion that it’s not clear that there shouldn’t be a property right here. Steve first says that “moving law in this direction is bad policy,” without providing any support for that assertion except to say that “Jerry has written extensively about the problems with propertization creep, so I don’t know why he would think that this makes sense.” I’ve already explained the difference between property rights that emerge at common law, and the creep we see from statutory property schemes, so I won’t repeat myself. The bottom line is that sometimes property makes sense, and sometimes it doesn’t. The question is, in a particular case, how do we increase human welfare? With a property right, or with a commons? As I said before, Steve and other commentators may be right that a commons is the way to go in this case, but it’s not as obvious to me as it seemingly is to them. I’d love to see the case made explicitly.

Steve goes on:

Jerry also turns to the economics of network effects to support his “it doesn’t feel right” hypothesis. As his argument goes, Craigslist built the network effects that it now enjoys, so competitors should have to do the same. I suppose that this satisfies a visceral sense of fairness, but it doesn’t say much about what is optimal for the market and for innovation.

I’d argue it says far more than Steve has said. As I explain in chapter one of Copyright Unbalanced (the subtitle of which Steve oh so cleverly zings in his concluding line) we do want to see new property rights emerge when it’s worthwhile to internalize (at least some of) an existing externality. In this case the externality is the network effect that listing platforms like Craigslist provide. What I’m suggesting is that it’s not obvious that we won’t get better platforms and more innovation by having competing closed platforms (Match.com vs. OKCupid) rather than one open one. Commentators like Steve, however, seem happy to forgo the analysis, dismissing such thinking as “Paleo-Schumpeterian,” and jumping right to the conclusion that any new propertization is bad. I’ve explained my limiting principle; I’d love to hear theirs.

Finally, so much of the reaction to Craigslist this past week was predicated on the company’s lack of innovation. As Steve points out,

Once you have a network effect in a market, your incentives to innovate decrease because of lock-in. Others, however, are strongly motivated to try to break into that market. Padmapper and others innovated—in a way that is no less “true” than Craigslist’s original innovation. Craigslist saw the value of that innovation and even tried to imitate it by creating its own mapping tool (arguably innovation in and of itself).

I don’t get it. If Craiglist is asleep at the switch and as un-innovative as Steve says it is, and if competitors are innovating, then why should we worry if it does have a property right? Won’t Craigslist eventually get disrupted and have the network effect wrested away from it by a competitor that entices aways its users? It’s how Facebook beat MySpace, and how MySpace beat Friendster. Lock-in didn’t stop them. And isn’t at least some of the incentive that draws those potential disruptors the chance of one day wearing the crown?

I’m not saying this is the right answer, but I don’t see anyone making a good case why a property model is obviously wrong, and a commons is obviously better.

  • http://www.facebook.com/nicmart Nicolas Martin

    This is disappointing, considering the source. It is poorly thought through and potentially leads to a new IP thicket in which to entangle American business.

  • sjschultze


    You aren’t trolling, are you? That’s the only reason I can think of for continuing to make such patently bad arguments, but you seem sincere. I guess I’ll portray your arguments in the most favorable light and presume that this is some kind of self-appointed pro bono devil’s advocate exercise (and no, I am not saying that Craig is the devil).

    I’m sorry that I brought up the copyright stuff. The problem is that coverage of this Order has almost uniformly asserted either that the copyright claims were demolished entirely (they were not) or that those that remain are unlikely to prevail (they very well could). In your post, you say that, “a site like PadMapper only copies facts about a listing.” Take a look at the First Amended Complaint, pages 11 and 12. That’s clearly not true. They’re not Feist facts. They’re not facts subject to the merger doctrine (although I commend you for incorporating another obscure if irrelevant corner of copyright jurisprudence into the discussion in this latest post. I’m sure that anyone still following along will find it educational.) In any event, even if your new merger theory were to prevail, the difference between the “facts” being non-copyrightable versus minimally enforceable under copyright law due to merger is splitting hairs. As for the validity of click-through contracts, I am sad to report that the caselaw here is very bad indeed for those wishing to to assert the tl;dr defense. Just ask Matthew Zeidenberg.

    Ok, on to things that don’t “sit well.” Let’s restrict the discussion to common law. What basis in common law do you think that Craigslist could plausibly assert? Apparently you don’t like trespass to chattels after all, even though you mentioned it in your first post. You don’t mention misappropriation here (also mentioned in your first post), so I guess that’s out too. There is nothing left, as pleaded. Apparently you’re proposing some new tort. Let’s call it the mis-scraping tort. You can’t scrape others’ web sites even if the content scraped consists of others’ (likely) copyrighted works because it doesn’t “sit well” because… well, because of the sweat of the brow? We don’t really like that line of reasoning here in America, but you may get somewhere with our European friends. In any case, it’s a novel intellectual property creation, and I doubt that many others share your “doesn’t sit well” opinion (besides, everybody knows that you have to write a Harvard Law Review article before something becomes a tort, so get to work!).

    On the economics of network effects. The benefits of network effects as they existed until these startups emerged were already being internalized by Craigslist. This yeilded lock-in or at the very least high switching costs and path dependence. This is common for winners in an industry, and it’s often bad for innovation because Schumpeterian creative destruction on the merits of the product is hampered. The Paleo-Schumpeterian argument is that the winner must always obtain maximum spoils (to the point of extended rent-seeking due to the aforementioned results of internalized network effects) in order to have the incentive to innovate in the first place. The obvious problem with this line of reasoning is that the “king of the mountain” position can become perpetually dominated by the schoolyard bully instead of the ingenious geek (a figure that some of us admire and identify with).

    “Won’t Craigslist eventually get disrupted and have the network effect wrested away from it by a competitor that entices aways its users? It’s how Facebook beat MySpace, and how MySpace beat Friendster. Lock-in didn’t stop them. And isn’t at least some of the incentive that draws those potential disruptors the chance of one day wearing the crown?”

    Don’t go Paelo on me, Jerry! Innovation can happen at many layers, and Schumpeter’s observation that disruption can happen in a given market is not a guarantee that it will. The Neo-Schumpeterian insight is that the waves of creative destruction are more like riddled basins of attraction (LSAT geeks in the house!) than sinusoidal curves. To put it in telecom terms, sometimes there truly is no “third pipe” on the way (oh yes, I went there).

    I didn’t argue that a commons approach is always the best for all information, or that a commons in classified ads is always best, but rather that the host of these classified ads should not be given (nor should it claim) some novel property right in them. Externalities should remain just that unless a compelling argument can be made that they should be propertized and internalized. Externalized externalities are good (because they benefit everyone) unless they dis-incentivize innovation on balance. That’s my limiting principle, and I think that the fact pattern here strongly supports the conclusion that more innovation would happen if this content is open rather than closed. I still don’t understand what your limiting principle is, beyond a gesture to Demsetzian “worthwhile”-ness.

    The online dating market is interesting, but it exhibits very different dynamics. The case for a commons there is weaker. This is in part because of the platforms’ true differentiation and their active role in adding value to user-provided content. It is also because the users’ incentives are aligned with their transfer of exclusive rights (which I imagine they do via click-through!).

    Back to footwear. We want a market favorable to new entrants, but where incumbents can make a buck two (or not, if they don’t care about that). I think it’s working fairly well for sandals. Crocs didn’t have to fend off a common-law property tort from Birkenstocks… or Jesus.

  • sjschultze

    Jerry, you’re obviously upset with my tone. You’re right that it came out as condescending. My poking was meant in jest, even though I do genuinely disagree with you here. We agree on far to many things to be enemies, and I like you as a person even if you think I’m a lout. I’ll be in Alexandria soon. Beer summit?

Previous post:

Next post: