Point - Counterpoint
How Do Not Track Can Save the Online Ad Industry
Note from the Editor:
This is the second post in a point/counterpoint between Mercatus Center Senior Research Fellow Adam Thierer and Center for Democracy & Technology Consumer Privacy Director Justin Brookman. Last week, Thierer explained why a Do-Not-Track standard will not have "a lasting, meaningful impact for long-term privacy protection."
This week, members of the World Wide Web Consortium’s Tracking Protection Working Group are meeting one last time in Sunnyvale, CA, to try to negotiate a Do-Not-Track web standard. Whether we succeed, and whether that standard is ever implemented by the online ecosystem, are very much in the air. However, the hope—my hope—is that the ad industry, browser makers and consumer advocates can come up with a consensus approach that lets ad networks deliver ads to support free online content while seriously curtailing the information they collect about users.
Last Thursday, Adam Thierer took a skeptical view of Do Not Track in this blog. Echoing his recent testimony before the Senate Commerce Committee, Adam cast doubts on Do Not Track as a silver-bullet solution to online privacy woes. He noted the previous failures of policy solutions (such as P3P and CAN-SPAM) to fix online privacy issues, and he posited that users might end up opting in to far more invasive tracking as a result.
By and large, I think Adam’s pessimism is misplaced. Certainly I agree with him that Do Not Track will not be a panacea for all—or even just online—privacy concerns. But it was never envisioned as a silver bullet. To the contrary, Do Not Track is rather humbly intended as a scalable way for users to tell the universe of third-party tracking companies to stop amassing profiles about what they do online.
For that reason, I remain perplexed that Do Not Track remains a controversial proposition in some quarters. Fundamentally, this is about self-regulation on a principle widely agreed upon: Users should be able to opt out of online tracking. Do Not Track largely mirrors what industry has already agreed to do by allowing users to opt out of behavioral advertising. Placing the Do-Not-Track instruction in the browser fixes two of the biggest limitations of the current industry approach—it makes the opt-out signal universal, so that everyone receives an opt-out signal, and persistent—so it doesn’t go away when you delete your cookies.
Historically, there have been disappointments, but we have also seen that technological policy solutions can have a demonstrable positive effect for consumers. The FTC’s hugely successful Do-Not-Call list is a good example; because of Do Not Call, we can opt out of telemarketers bombarding us with unwanted solicitations. The Fair Credit Reporting Act gives us access to our credit reports and lets us contest errors that compromise our ability to get credit or a job. Do Not Track may in some ways be less ambitious than those efforts, but it represents a way to fulfill a preference of (some, not all) consumers to stop unknown companies from watching them across the web.
As to the contention that users will agree to more invasive tracking as a result of Do Not Track, it is not my intention to dictate market outcomes. If users are content to agree to be tracked in exchange for content, so be it. Do Not Track is merely intended as a signal to sites that certain users don’t want to be tracked around the web. If sites want to countermand such an instruction, let them make a pitch to their customers. I personally doubt that a lot of publishers will require cross-site tracking in exchange for content, but it’s certainly possible. Do Not Track simply allows for an expression of intent about an industry practice that has existed in the dark for a long time. The market will respond to that signal one way or another. We should not paternalistically prevent users from expressing a preference not to be tracked because we feel the end result will be better for them.
There remain hard questions for the Working Group to resolve—most notably how much data should be collected even when Do Not Track is turned on. Can data be collected for research purposes when Do Not Track is set? Can ad networks use unique identifiers like cookies despite the signal? Fortunately, in recent days we have seen more willingness on the part of industry to address these issues within the Working Group.
This renewed commitment to negotiate is undoubtedly tied to stronger actions by the browsers to combat ever-increasing online tracking on behalf of their users—most notably Mozilla’s decision to join Apple to block tracking companies from setting cookies. If the W3C is unable to come up with a consensus Do-Not-Track standard, these retaliatory efforts are certain to continue.
Ultimately, if industry cannot agree to a model that allows for privacy-preserving ads, they may be prevented from serving ads altogether. Do Not Track represents an effort to set the basic rules for engagement in response to a user’s preference. Industry alone has made substantial strides in recent years, but to date those efforts have not been sufficient to alleviate the concerns of advocates, regulators or consumers. Hopefully this collaborative approach can forestall further escalation and allow the model of ad-supported online content to persevere. I, for one, remain optimistic, as the alternatives may be worse for all concerned.
About the Author
Justin Brookman is the Director of Consumer Privacy at the Center for Democracy & Technology (CDT). Prior to joining the CDT in January 2010, Brookman was chief of the Internet Bureau of the New York Attorney General's Office. Under his leadership, the Internet Bureau was one of the most active and aggressive law enforcement groups working on Internet issues, and Brookman brought several groundbreaking cases to protect the rights of online consumers. Brookman previously worked as a litigation associate for six years at Fried, Frank, Harris, Shriver & Jacobson LLP in both its New York and Washington offices.