When progress trumps privacy
Baker’s argument has been panned in civil libertarian circles. When he suggests that societies eventually adapt to new technologies — that “the raw spot grows callous” as we grow accustomed to invasions of privacy — privacy campaigners reply that it is Baker who has grown callous to the harms in question. Baker’s central goal is to convince Americans to accept that government must use new technological tools, like the data mining programs used by the National Security Agency, to combat mass-casualty terrorism. His critics maintain that he is far too glib about the potential that government might abuse these new tools, and indeed too dismissive of the notion that it has already done so.
I’m torn on the question of whether the national security state has overstepped its bounds, and there are people I respect on both sides of the debate. Civil libertarians like Ben Wizner of the American Civil Liberties Union and Julian Sanchez of the Cato Institute see the new Leahy-Sensenbrenner USA FREEDOM Act– which would end the dragnet collection of Americans’ phone records under the PATRIOT Act, and limit other surveillance — as an important step towards reining in a bureaucracy run amok. Baker fears that it will cripple the ability of U.S. intelligence officials to prevent future terror attacks. I couldn’t tell you which side is closer to the mark.
What is increasingly clear to me, however, is that privacy concerns are limiting our ability to flourish as a society for reasons having nothing to do with NSA surveillance.
The Food and Drug Administration recently ordered one of America’s most popular consumer genonics firms, 23andMe, to cease selling and marketing its direct-to-consumer DNA test on the grounds that it is a medical device subject to FDA approval. The FDA’s case seems pretty flimsy. The saliva collection kit that 23andMe offers through its Personal Genome Service is utterly harmless, and no one is claiming otherwise. Rather, the FDA is concerned that by giving its consumers data on disease risks, complete with plenty of disclaimers, it may prompt them to seek unnecessary MRIs and mastectomies, as Christina Farr of VentureBeat reports. The obvious rejoinder to these concerns is that consumers don’t have the option, for better or for worse, of operating on themselves. They generally need a medical practitioner to sign off, and medical practitioners hardly suffer from a lack of licensing and regulation. The FDA seems to be engaging in a senseless power grab.
The reason this matters is that 23andMe represents just the first step of the coming consumer genomics revolution. Recently, Razib Khan and David Mittelman outlined the future of consumer genomics in a short article in Genome Biology. First, Khan and Mittelman expect that startups like 23andMe, which offer consumers basic information about their ancestry and genealogy, to grow more popular as the costs of their DNA tests continue to plummet. Then these vast databases will be used to yield real scientific insights, as biomarkers record how we respond to the food we eat and the activities we undertake, and this data “is intersected with millions with varying levels of genetic relatedness and lifestyle.” The result will be “a perpetual stream of novel insightful predictions,” and Khan and Mittelman see this future as all but inevitable. Yet for this future to become a reality, consumers will have to grow more comfortable with sharing their personal medical data. One of the reasons Americans are so sensitive about sharing this data is that many of them fear becoming uninsurable, a fear that universal coverage will (hopefully) do much to allay. If privacy concerns win the day, the marriage of big data and personal genomics might never come to pass — and our best hope for achieving medical breakthroughs in the decades to come will be dashed.
It’s not just a desire for medical privacy that’s getting in the way of progress. U.S. higher education institutions have fought tooth and nail against efforts to build a unified database of “student unit records” — collected throughout a student’s educational life, and anonymized – that can allow taxpayers, parents and students to see how different kinds of students have fared at different colleges and universities. As Kevin Carey, director of education policy at the New America Foundation, has argued, higher education institutions understand that if this data is released, the federal government can hold them accountable for their performance. For example, a student unit record system will reveal which schools do the best and worst job of educating low-income students who receive Pell Grants. This is a prospect that keeps higher education administrators up at night, and with good reason, as it threatens the flow of federal dollars to subpar educational programs.
So how do the higher education institutions get away with keeping Americans in the dark about how well they are educating U.S. students? It’s simple. Critics of a federal student unit record system warn that it represents a threat to student privacy — determined sleuths might be able to figure out the grades and household income levels of individual students, despite efforts to anonymize the data. And though there are many techniques schools and governments can use to protect privacy, it really is true that data anonymization is a hard problem to solve. The question is whether we should put privacy ahead of the goal of building a more efficient and equitable higher education system.
Privacy matters. But so do the things we give up when we let the fear of invasions of privacy stymie the development of promising new technologies.