TOC The tyranny of the freedom of speech Your Turn

December 10, 2008

Three weeks already since Stephanie Saul reported (*) a decision by "a federal appeals court in Boston" upholding the constitutionality of a New Hampshire law which forbids pharmacists to hand over physician prescription data to drug companies (1). I alluded to this law two years and a half ago in a fillip on doctor practice profiling. Privacy was at stake without the high drama inherent in patients' violations. If I expressed some doubts then concerning the enforcement of such laws, I can only rejoice now in learning of the Boston decision.

Eprivacy is far from winning the war though. This victory is less than it seems as privacy played a minor role in the judges' opinion (**). To justify the outcome, a judge is entitled to play his or her strongest card, in the instance the right of New Hampshire to "advance its substancial interest in reducing overall health care costs". But relative legal weakness bodes ill for the health of privacy protection.

The reality is dire indeed. Data aggregators acting on behalf of drug companies, IMS Health and Verispan had attacked the New Hampshire law as curtailing freedom of speech. Were he touring the US today, this would have left Alexis de Tocqueville speechless. But writing for the majority, Circuit Judge Selya declares that "[while] scholars have labored to formulate theories about why First Amendment immunity exists in such cases [...] The matter remains a doctrinal mystery". Tocqueville would have warned us about a tyranny of the freedom of speech.

How can privacy prevail then? Before we look at Judge Selya's arguments, let us tackle his colleague's concurring dissent. For Judge Lipez, the law does limit freedom of speech but still complies with the US constitution as its restrictions concern so called commercial speech and arise from a legitimate government interest. So if tomorrow the State of New Hampshire saw a way to cut costs by buying physician prescription data from IMS Health, Judge Lipez would approve as well. For eprivacy, such an ally is purely tactical. Long term he is downright dangerous.

Judge Selya's approach is far better and shares some kindred spirit with the work of Professor Neil M. Richards, one of the scholars he mentioned. While these two lawyers may object to be so summarily lumped together, I will avail myself of my freedom of speech rights and say they both take a reductionist line. Dealing with targeted advertising, Judge Selya divides the behaviors under consideration and finds the law affects not speech, i.e. the underlying information flow, but conduct, i.e. the use to which this information is put to. In one paper (***), Prof Richards starts by dividing the tasks into "collection", "use", "disclosure" and "marketing" (2) and point by point shows how to "resist[...] the creep of First Amendment analysis".

I dare not delve deeper into the details of this "divide and conquer" strategy. While it provides useful ammunition in defense of privacy, I believe it suffers from a potentially fatal flaw. Opponents will argue targeted advertising deserves to be judged as a whole and, following Portia, assert that governement can spare free speech when ruling on targeted advertising no more than a knife can spill no blood when plunged into the flesh.

Indeed Professor Richards has come up with a new, clever solution (****). Instead of trimming the powers of the First Amendment, he now yields it as a shield to protect freedom of thought. Together with "spatial privacy, the freedom of intellectual exploration, and confidentiality of communication", freedom of thought makes up "intellectual privacy", equal in dignity to freedom of speech under the US constitution.

As much as I find his argument persuasive, I feel that, by excluding "purchases of consumer goods" as "fundamentally different in kind" from checking a book from a library or searching by keyword on the Internet, he leaves too much to fortune. An obvious truth when the consumer good is itself a book, the purchase of all goods and services can be, as marketing managers well know, branded into an "exploration" beyond the satisfaction of a purely material need. Buying an Apple computer may rank lower on the scale of intellectual activities than looking up "Democracy in America" (3). Yet do not hoi polloi deserve privacy protection as much as the happy few?

My own theory of eprivacy is to grant title to their personal data to all, under the protection of the Fourth and Fifth Amendments. Reading the learned authorities quoted above, I now perceive two serious weaknesses. Not all rights in the US constitution seem to have been created equal and, like the French Consuls (4), the First reigns supreme. There is a further risk. As the US bill of rights protects against mischief from the government, invoking the Fourth and the Fifth does nothing to deter privacy violations from health data aggregators among others.

Some US judges would rather die than quote foreign law, especially when New Hampshire is concerned. Free from such myopia, I seek protection in the French Declaration of the Rights of Man and the Citizen, which preceded the adoption of the US Bill of Rights (5). There the rights to property and resistance to oppression (article 2) have priority over the freedom of thought and speech (articles 10 and 11) (6).

If data privacy are but property rights, it remains to be seen how they can be retained by the original owners. Possession of course is not to be confused with ownership. Although I may find myself defrauded, I do not surrender all legal rights when I deposit my savings in a bank. Be it data or money, rules readily limit how one may use what it receives from another party, a notion amply illustrated by both Judge Selya and Professor Richards (7). The right of property simply turns "the requirement that data collected for one purpose may be used for that purpose only, absent consent" into a principle, opposable to all, private or public. It limits the freedom of speech of those in possession of my data no more than my inability to advertise in the New York Times for lack of the proper resources.

For instance if I search for "eprivacy" and absent my consent, all records of my search ought to disappear as soon as my search has been answered. Keeping them could serve no purpose and would amount to reckless endangerment. For longer lasting transactions such as purchases, records would become confidential by default, limited to supporting payment and warranty. Notice that nothing prevents the record holder to keep its own accounting and statistics. Whether "privacy" is a popular keyword or "Apple Iphone" is a best seller is the property of the provider, not the user's.

Consent can of course be granted and, by doing so, we step down from civil rights to commercial rights. Professor Richards would recognize there what he deems "the basic dualism of modern constitutionalism". He may approve the scheme but will sadly point out that current privacy policies are precisely the contracts which detail to what users consent. To avoid such an abject surrender, am I then asking the government to step into such a morass? This is how rules get a bad reputation. My approach is simpler.

I recall my three legal informants emphasize the need to "level the playing field", in Judge Selya's words. Judge Lipez speaks against "undue influence". Reviewing history, Professor Richards recalls how labor reformers sought to eliminate "unequal bargaining power by enacting social legislation". Invoking the universal right of resistance to oppression in the same spirit, I suggest not a rule, but a law which makes it illegal to require a user to consent to any "secondary use" of his or her data as part of any transaction whose sole object is not to collect user data.

Bundling is what tilts the playing field against individuals. Users have no need to be protected from Nielsen by the government. Companies will say the money they get from violating user privacy is needed to balance their budget. Perhaps, but why not let competition find a better business model? De facto monopolies may threaten to close their service altogether. Why would Google ask users to tell Congress they would rather live than be free? Its original business formula does not rest on privacy violations and there are laws to penalize such hypothetical abuse of market dominance.

Freedom of speech is not there to protect Google's rights to target advertising. Rather it is Google which ought to protect First Amendment values. As Jeffrey Rosen analyses in depth, the extent to which Google censors content creators to appease local governments is a real challenge (*****). Jude Webber and Rob Minto tells how such censorship requests come from prominent Argentinans (******). Notice Google is neither a truly blind recommendation algorithm nor a normal mass media passing editorial judgment. The resulting arbitrariness is a form of speech and ought to be free.

When, through market dominance or otherwise, the playing field is not level though, unfettered commercial speech is never far from tyranny.

Philippe Coueignoux

December 2008
Copyright © 2008 ePrio Inc. All rights reserved.