Search Law & Policy @ Haifa

Search is becoming an increasingly important topic in cyberlaw. After several sessions and conferences looking at the issue (including at CFP 2004, Yale Law School and in Berlin), Niva Elkin-Koren and Michael Birnhack from the University of Haifa in Israel have now put together a great conference on The Law of Search Engines, which I am currently attending. I won't do a full blogging of the conference, but I just wanted to make a few notes of what I found particularly interesting:

  • Alluding to Yochai Benkler's work, Niva Elkin-Koren and Sheizaf Rafaeli talked about the problem that search engines such as Google killed the idea of disintermediation and reintroduced an intermediary into Internet communications. This is not only problematic if one is interested in how the Internet alters public discourse, but also because it creates new targets for regulation for purposes of censorship, content control, tax purposes etc. This point led to a discussion about to what extent disintermediated public spheres on the Internet can be imagined, which also relates to Cass Sunstein's Infotopia.
  • Michael Geist pointed out that the identity of intermediaries may change. When, e.g., Wikipedia appears in the top three ranks of Google results for a wide variety of search queries, it is actually Wikipedia, not Google, which is the relevant intermediary.
  • Bracha Shapira presented MarCol, an interesting collaborative research project that attempts to incorporate the evaluation of search results by other users into the selection of search results.
  • Helen Nissenbaum presented TrackMeNot, a Firefox extension that attempts to protect the privacy of search engine users by "hiding" actual search queries into a large noise of bogus queries. It's basically privacy through obfuscation applied to search engines. This created quite some discussion about whether such obfuscation techniques create an undiserable "pollution" of the information environment or whether they are legitimate tools for users to combat privacy erosion. This is just one example of a broader debate which can be observed in many areas of the law (e.g. patent applications, mass market contracts, code obfuscation technologies to prevent reverse engineering, even the analysis of Ehud Kamar concerning Delaware's strategy in corporate law, etc.). In general, the legal system does not seem to have developed good mechanisms to deal with the conscious creation of information overloads.
  • Tal Zarsky gave an interesting talk about search engines and media policy and urged scholars and policymakers to look at empirical evidence about issues such as substitute relationships before engaging in a media law analysis. In his talk about search engines from an antitrust perspective, Avishalom Tor pointed to problems of applying antitrust laws to search enginges (such as the proper market definition when the product is, at least apparently, offered for free). He also pointed to limitations of regulating search engines by antitrust law, including the speed of (or lack thereof) antitrust regulation. Michal Gal seconded by pointing to other limitations such as the problems competition policy has in addressing coordinated effects in oligopolistic markets, markets with strong information asymmetries or markets in which consumers could inform themselves about certain product dimensions, but simply do not care about them (as it is sometimes experienced in the privacy context). In general, antitrust can only be seen as part of a larger toolbox that tries to address policy problems created by search engines.
  • Michael Geist talked about unintended consequences of the search age. Pointing to David Brin's Transparent Society, he showed how data that once was thought private is increasingly becoming publicly available. In addition, he pointed out that it is not only the search database itself which can be scary. Analyzing search queries can also reveal substantial information about the user issuing the query, thereby creating a "database of intentions". (And here is just a link to a CfP if you are interested in query log analysis).


The reason why Wikipedia shows up so much for searches is because they have a lot of content and are also very good at search engine optimization. The philosophy is quite simple. Wikipedia makes a lot of search pages, and optimizes them well so users can find them and link to them.

Add new comment