More on Google: is scraping anticompetitive?
One of the good things about my powerpoint presentation on “Search engines and competition law” is that it does not give a clue about the substance of what I said 😉
As you can infer from my previous posts on this investigation, I’m quite skeptical (to put it mildly) with regard to the legal basis for the accusations against Google (partly because I don’t know whether there are any…).
Whereas in my presentation I mostly addressed search bias-related allegations, there is another issue that raises pretty fundamental questions. I’m referring to the claim that Google “scrapes” content from other websites and appropiates it (these allegations refer to the small portion of black letter text that contextualizes search results and that you find below the green web address shown below every result).
Allegations over this practice strike me for various reasons and raise a few important questions, mainly what on earth does it have to do with competition law?
If the content at issue is not copyright-protected; if they are short informative excerpts originating from content made publicly available; if they contribute to “scraped” pages getting more traffic; and if they’re necessary for users to understand search results; why should competition law ban their use? Why should competition law create a second level of protection stricter than copyright law? Why second-guess copyright law? This brings us back to an old debate: where does competition law stop? Should antitrust also second guess, for instance, environmental law, tax law, accountancy or data privacy? (I’ll come back to the data privacy-antitrust interface on a later post, for it “coincidentally” appears to be a hot academic topic nowadays).
Allegations on scraping were discussed within the framework of the FTC’s investigation. In fact, as part of the settlement with the FTC Google has committed to allow all websites to opt out of appearing on Google’s vertical sites while still appearing in organic results.
Given that complainants in the EU proceedings have also insisted on this matter [(probably scraping content from each others’ complaints ;)] it’s safe to guess that the European Commission might attempt to obtain a similar commitment, and I find this quite a problematic precedent. Againg: why is scraping a competition law issue?
I’m comforted not to be the only dissenter. Commissioner Rosch authored a dissenting opinion accompanying the FTC’s statement in which he articulates the view that these practices “do not violate the antitrust laws in any respect“. He notes that his colleagues manifested “strong concerns” about these practices, but that no one ever argued that they breached antitrust rules, not even Section 5 of the FTC Act.
For a sensible opinion on why the display of content from other websites in search results is a perfectly valid business practice, take a look at the German Supreme Court’s Judgment in the Paperboy case (click here for a summary in English)
Chillin'Competition 
Competition law by its very essence “second guesses” other fields of law: it finds an agreement,where contract law would never find,it classifies otherwise perfectly legitimate capital increases and tax exemptions as state aid,and yes,it may consider a conduct as an abuse if it is fully legal under all other concievable fields of law (think of vexatious litigation or the application of very low prices as an example…) competiton law has a different aim from other laws and if the protection of consumers requires a certain otherwise lawful practice to be stopped,then be it!
Asimo
14 February 2013 at 11:32 pm
Very right, competition law is about second guessing. But this is only to the extent that the conduct at hand has anticompetitive effects (i.e. foreclosure). And I have to concurr with Alfonso that such types of effects are far from obvious here.
Nicolas Petit
15 February 2013 at 11:23 am
Just found this on the web, another good presentation on market definition in search: http://www.agcm.it/trasp-statistiche/doc_download/3174-convegno20120524zingales.html
Nicolas Petit
15 February 2013 at 12:09 pm
Asimo, thanks for an excellent comment. I’m particularly grateful because you got Nicolas to agree with me, which is a remarkable achievement 😉
On the substance: you make a very sensible point. In fact, by making the post somehow provocative I may have made it loose technical rigour. The point I’m trying to make is (i) that the use of the content at issue cannot constitute “misappropiation” under IP law, and (ii) that its use is objectively necessary for search engines to provide their services; that it is useful for consumers; and that, moreover, it enables “scraped” sites to receive more traffic.
Under those circumstances, I’m assuming that “scraping” is a legitimate/objectively justified business practice (competition on the merits) that can in no way lead to foreclosure. That is why I presume that arguments against these practices are grounded on some idea of “unfairness”, and this is what I see as problematic.
Alfonso Lamadrid
15 February 2013 at 2:48 pm
I don’t see it either…From IP and unfair competition law perspectives, see end of Ryanair’s battle in Spain against online travel agencies for scraping: Spanisn Supreme Court Ruling STS 572/2012
Ainhoa
15 February 2013 at 2:56 pm
I have no choice but to jump in!
Alfonso, I think I agree with your post more than you seem to do.
The fact that competition law overlaps with other disciplines does not mean that it second-guesses them. It is not because competition law applies alongside contract law, or because some tax measures qualify as State aid, that the competition law system is second-guessing those areas of the law.
Alfonso’s point is finer and more sophisticated, and one with which I fully agree. These scraping claims are an invitation to re-invent copyright law and strike a different balance for the protection of information. And as such, dubious.
Bagnole
15 February 2013 at 3:26 pm
I fully agree: the question is – as both Alfonso and Nicolas has pointed out – whether there are (proven) anti-competitive effects on the relevant market (e,g, foreclosre), ie whether the practice harms competition/consumersors in order for such practice to be qualified as a breach.
The point was that it is not necessary (and almost irrelevant) to check whether something is legal or illegal under IP law (or other fields of law) and this can even be misleading as it takes our eyes away from being able to respond to the above fundamental question of the harm to consumers.
Asimo
15 February 2013 at 4:34 pm
Asimo, unfortunately (in my view) that is not the question the law poses. Article 102 jurisprudence does not require “(proven) anti-competitive effects on the relevant market e.g foreclosure)” as you state the evaluation to be. I quote from AstraZeneca (citing TeliaSonera) (para 112):
“[A]lthough the practice of an undertaking in a dominant position cannot be characterised as abusive in the absence of any anti-competitive effects on the market, such an effect does not necessarily have to be concrete, and it is sufficient to demonstrate that there is a potential anti-competitive effect.”
That is a markedly lower threshold. While, to me, it is quite clear that Google’s scraping is not anti-competitive, the threshold for that evaluation is set absurdly low by the Court as noted above. No wonder Google is attempting to settle.
Colm
15 February 2013 at 5:33 pm
This is not the point Alfonso was trying to make, which goes to the essence of competition law systems.
If the information used by Google is freely available, in the sense that it is not protected by copyright, there cannot be anything anticompetitive in using it.
Competition law is there to protect the competitive process, not to perpetuate particular companies or business models. It is in the essence of the competitive process that some firms leave the market. It does not necessarily mean that it is the result of anticompetitive conduct. It may be the case that their business model is obsolete, or that it does not offer an added value relative to, say, Google’s services.
Now, if you want to protect particular firms of business models by adding an additional layer for the protection of informations, what you are effectively doing is re-creating a system of intellectual property protection, not protecting the competitive process.
Bagnole
15 February 2013 at 6:04 pm
Colm, point well taken – my comment was based on my view on the the ideal state of competition law in general as opposed to current EU competition law as interpreted by the CJEU.
Bagnole, I think it is now a competition law commonplace that “competition is is to protect consumers / the competitive process and not competitiors”. However, the reality is that it is very difficult to make a difference between the two in individual cases, as the harm to a competitor could very well be a very important/telling sign (but certainly not a final proof) of a failure of the competitive process in a given relevant market. At the same time, your general statement that “If the information used by Google is freely available, in the sense that it is not protected by copyright, there cannot be anything anticompetitive in using it.” were true, then we would never have illegal exchanges of public information (see the Horizontal Guidelines of the Commission)…
Asimo
15 February 2013 at 7:42 pm
Asimo, you may want to re-read the Guidelines on horizontal co-operation agreements to see the legal status of exchanges of publicly available information (para. 92). The idea underlying the Guidelines is exactly the same: if the information is publicly available, there cannot be an infringement of competition law (the competitive process cannot be said to be altered in any way as a result of the exchange).
You should also bear in mind the difference between unilateral and coordinated conduct. This is the crucial factor underlying Alfonso’s point. If Google displays (unilaterally) information that is in the public domain, and nothing else than that, what is exactly the anticompetitive behaviour? Conversely, if the information is not in the public domain, then this is a matter for copyright.
Many firms go bust every day because they are unable to cope with competition. This is uneventful and a normal aspect of business life. It has nothing to do with the objectives of competition law and the (yawns) protecting competition vs. protecting competitors debate. As such, the fact that a firm has to leave the market is not an indicator that anticompetitive conduct has taken place. One has to go the other way, and establish anticompetitive conduct in the first place. Which brings me to the previous paragraph. Why exactly would the naked use of publicly available information be anticompetitive?
Bagnole
15 February 2013 at 8:33 pm
Bagnole,I beg to differ:
– the horizontal guidelines – i suggest to go back to para 83 – clearly confirm that even the exchange of fully public info can be problematic and that there can be an infringement (“Even if there is public availability of data (e.g., information published by regulators), the existence of an additional information exchange by competitors may restrict competition. In this case, it is the incremental information that is critical to tip the market balance towards a collusive outcome.”) . I do not say this is the case in google,i just merely wished to refute the general allegation you made in this repect.
– “if the info is not in public domain,then the matter is for copyright”.Just one word:Magill. That aggregated TV guide would never have been produced if the ECJ only had relied on copyright law.This was the exact point I wished to make to Alfonso: competion law has different aims from copyright and the same conduct may have different reprecussions in the two regimes.
– surely firms enter and leave markets all the time,but competiion law is precisely about assessing these events and deciding wheter these are a result of “normal” market flows or perhaps evidence of harmful market structures (barriers to exit/entry) and behaviour (such as predatory pricing,etc.).unless you ab ovo deny that exclusionary conduct could be abusive…:)
Asimo
15 February 2013 at 11:41 pm
While I agree with the overall sentiment, it might be helpful to consider a more specific factual allegation. As I understand it, one of the most vociferous complainers (in the U.S. anyway) about “scraping” was Yelp, which was unhappy that user-generated reviews from it’s site were showing up in Google Places pages.
If one searched, for example, for “restaurants near Avenue d’Auderghem, Brussels”, Google’s “one box” or places results allegedly would show comments about, for example, Park Side Brasserie that had originally been left on Yelp. (Doing that today does not lead to this results, but I’m fairly sure it used to).
From Yelp’s point of view that is quite troubling, as that user-generated content is its entire business. If users don’t need to go to Yelp to get it, there might be a potential for foreclosure there.
I think it’s still hard to determine whether the foreclosure effect for Yelp is greater than the advertising effect (assuming those reviews are labelled as coming from Yelp), but the it’s a trickier fact pattern than summary content that appears under search results.
Adam Miller
15 February 2013 at 8:17 pm
[…] Alfonso Lamadrid questions whether Google’s alleged “scraping” of publicly available information has any anticompetitive effect. […]
Antitrust Links of the Week « Uncategorized « UpwardPricingPressure
15 February 2013 at 9:26 pm
@Asimo
May I suggest that you train your reading skills and that you pay particular attention to the words ‘additional’ and ‘incremental’. Attention to detail is crucial in this craft!
For the rest, nobody is denying that exclusionary conduct can be abusive. The trick here is that it is not clear what the ‘conduct’ is in the alleged scraping allegations. How can using information that is in the public domain abusive? How on earth can one argue that this is that not a form of competition on the merits? These are the preliminary questions that cannot be avoided/
Bagnole
16 February 2013 at 12:14 am
Bagnole:
Without wishing to provoke a virtual swordfight here:))) just to clarify: the point regarding the guidelines is that there is (i) public info already available,such as a regulators publication on a market or a website showing “user generated reviews” and (ii) there is an additional conduct by competitors,such as the exchange of that public info or a scraping of the same info on your website.the analogy is certainly false to the extent that in google we talk about unilateral and not coordinated behaviour,however,the point is the very same: that fact that your conduct (part ii) relates to info already in the public domain (part i) does not mean that you are automatically safe from antitrust scrutiny.
Asimo
16 February 2013 at 7:18 am
[…] ‘scraping’ concerns raised in the context of the Google investigation provide an excellent example. The […]
NEW PAPER: Restrictions on Innovation in EU Competition Law (and other news!) | Chillin'Competition
21 December 2015 at 9:16 pm
[…] Article 102 TFEU, including data protection (excessive data collection), intellectual property (content ‘scraping’), consumer protection (opaque and arbitrary application of operating rules) and unfair […]
Op-Ed: “The Missing Link in Article 102 TFEU: Defining the Boundaries of Competition Law Enforcement” by David Pérez de Lamo - EU Law Live
12 January 2021 at 8:44 am