(Jeffrey Rosen is a Professor at George Washington University School of Law and author of “The Naked Crowd).
ROSENZWEIG: How long have you been examining – the broad topic of this project is government databases, data mining, link analysis, government surveillance, and collection of information. How long have you been working in that area?
ROSEN: I’ve been working in privacy generally since the late ’90s. My first book on privacy was published in 2000, but I focused more on government data collection ever since 9/11, obviously, as many people did, and started writing about it beginning right after 9/11. And my first book on government data gathering was published in 2004, and that was “The Naked Crowd.”
ROSENZWEIG: Do you find writing about data mining different from writing about other subjects? And if so, why?
ROSEN: It’s a tough thing to write about. It’s tough to write about mostly because it’s hard to come up with identifiable victims. And as you know, privacy is easiest to write about when you can give concrete examples and anecdotes about the harms. When I write about data mining, I’m also driven first to talk about the Nixon effect. It’s the fact that President Nixon during the Vietnam era wanted to go after his critics and look for improprieties in their tax returns. But the Nixon effect doesn’t resonate with majorities in a very dramatic way. Civil libertarians and libertarian conservatives are worried about it, but there’s a broad feeling that if you’ve done nothing wrong, you’ve got nothing to hide. And the idea that the government shouldn’t look broadly for evidence of terrorism and then prosecute people for low-level crimes, which is another problem with data mining, the Al Capone effect, doesn’t resonate for the same reason as the Nixon effect. So, the combination of the fact that the harms are often elusive and that the actual procedure of data mining is technical and that it doesn’t affect people in tangible ways, the way, for example, body scanners at the airport affects people, makes it a hard thing to write about in large popular venues.
ROSENZWEIG: And yet, when you wrote “The Naked Crowd,” which is essentially about things that affect people far more tangibly, or at least a portion of it, even there you found people disregarding or diminishing the privacy-invasive nature of what was happening to them. What do you think explains that? Or have you got a theory on how that comes about?
ROSEN: For me it’s a dramatic example of a harm that’s easy to make concrete and a surprising policy result that I didn’t anticipate when I started writing about it. It might be useful for data mining.
So, broadly, when I wrote about the body scanners in 2004, it was obvious even then that the government was being presented with a choice between a naked machine and a blob machine. In other words, a body scanner that showed the human body naked and a body scanner that was just as effective from a security point of view but that scrambled the body into a nondescript blob. And the people at the Pacific Northwest Laboratories presented the government with that choice. You would have thought that it was a no-brainer. The one machine threatened privacy. The other promised just as much privacy – or rather, just as much security without threatening privacy, a sort of silver bullet technology.
But the Homeland Security Department rolled out the naked body scanners without public comment and without questioning their privacy-invasive potential, and then appeared surprised by that backlash. I also was surprised by the result of the backlash, although the protests against the body scanners were dismissed by some as having fizzled, that immortal protest by the computer programmer, “Don’t touch my junk,” galvanized enough of a public reaction that the Homeland Security Department just a month or two ago backtracked and said that it would in fact try on a pilot bases bloblike screening technologies at Reagan and a couple of other airports and perhaps use it more generally.
So, for me, the fact that it took seven years for the blob technology to be reluctantly adopted by the U.S. government under pressure, and also is striking, but it mattered that people personally and tangibly experienced the indignities of the body scanners at airports, the fact that millions of people had to go through them was enough of a minor indignity and annoyance, that at least vocal minorities were mobilized into protests, and that prompted a response. So, that suggests that data mining may be also difficult to mobilize people against unless travelers at airports, for example, on a large scale, experience the indignity of being excluded or put into different lines or classified somehow on the basis of data profiles. It takes something that’s tangible as people grabbing your private parts to get people mobilized about the effect of some of these technologies. Without that effect, I wonder how easy it is to mobilize citizens.
ROSENZWEIG: Do you find that attitudes toward things like data mining and privacy vary based on any demographic, like age or gender or country of residence?
ROSEN: You know, yes, indeed. There are demographic and age differences. It’s commonplace that young people are said not to care as much about privacy as older people, and that Europeans are said to care much more about privacy invasions by the private sector as opposed to Americans, who are more concerned about privacy invasions by government.
But all of these generalizations break down depending on how you ask the questions. There have been good studies by Pew suggesting that younger people do, in fact, care about cleaning up their Google profiles when they start losing jobs on the basis of it. It’s really important not to talk about privacy in the abstract because people have different baselines of expectations and define privacy in different ways. I’ve found, speaking very broadly, that besides the body scanners, the privacy issue that’s really galvanized most people since I’ve been writing since 9/11, I would say it’s a problem of digital forgetting on Google that the drunken Facebook pictures that are losing – that they’re causing people to lose jobs and promotions is something that everyone has experienced or knows someone who has. And that is something that young people definitely care about as much as older people, although perhaps in different ways. And that’s something that you can write about and speak about and get a tangible and visceral response.
ROSENZWEIG: So, let’s transition a little bit. We’ve been doing theory. Now I want to ask you kind of about how to do reporting and what you do to get stories that are in this area. So, let me start broadly and just invite you to tell me about some of your experiences in trying to find information and stories about data mining and surveillance and link analysis, and what makes that hard or easy in how you go about it.
ROSEN: For me, the toughest thing is finding a topic and framing it in a way that will interest editors, in particular the editors at The New York Times Magazine, where I’ve been writing mostly about this subject. Certainly, the policy community is completely open to you and technologists are happy to help. In terms of getting access to the government, it helps to have an assignment from the Times Magazine, for example. And when I have, I’ve had no trouble talking to anyone I want within the government or the FBI or Homeland Security. When I was writing for the Atlantic initially about Michael Chertoff, I sat down with him a lot. So, I would say that access depends often on the venue that you’re writing in. But many of these topics don’t require a huge degree of access. So, the real challenge is formulating a topic and selling editors on it.
ROSENZWEIG: OK. And then what would you recommend in terms of trying to do that? Is it topics that affect people who are the readers or is it topics that hit on new technology? What types of topics that if well formulated generically you think push the story well?
ROSEN: My first piece I did for the Times Magazine right after 9/11 was about surveillance cameras in Britain combined with the effort to do some linkages in the data space. And that was just an unusual fortuity in timing. I’d gone over to England weeks before 9/11 to write about surveillance cameras, which I just pitched as a straightforward privacy story, and then after 9/11, that obviously became much more salient, and I wrote it very quickly under big deadline pressure. It came out in the weeks right after 9/11.
And then since then, there have been a series of stories – let me see what the other ones have been related to this. Because the digital forgetting story about the private Facebook pictures I mentioned was last June, and that was easy to pitch because it was of such broad interest in people in the age of Facebook. It could have a national security component.
There have been privacy-related technology stories, one on neuro law and how brain scan technology is transforming the legal system, which, you know, is related to this broadly. And that was another case where it was just a huge foment in the academy and technology was changing a whole area of law and across a whole range of doctrines. So, once I convinced them it was intellectually and journalistically interesting, they bought that.
But in most of these pieces, I’ve found that there’s a series of discussions over a period of months about what will be broadly interesting, and then some access to pin it down, and then you get the pitch accepted. The piece I’m working on right now is in the same vein. It’s about how real-time bidding and targeted Internet ads are transforming online experiences in journalism. And here again, although it’s not a national security piece, understanding the technology of these targeted ads and figuring out the social implications takes a long time.
So, basically, my experience, which has been very satisfying, in doing these pieces, is just that you’ve got to really work hard at the front end to figure out a formulation and a topic that will appeal to lots of people and have to learn a lot about it and learn enough about it that you can make it compelling.
ROSENZWEIG: Do you think people who want to write in this area need to learn some technology background, or do you think it’s something they can learn as they do?
ROSEN: I think you can learn as you do. I certainly am basically a neo-Luddite. I’m not all that sophisticated. I might be a quicker study if I had more and if there were courses that were easily available or if the resources were there, I’d certainly be delighted. You know, that would be useful, and I’d love to look at them. But you certainly don’t have to be especially tech savvy to write about this stuff. Or in law. I mean, it helps to have some expertise or familiarity, I guess, but people who write about this a lot all have different strengths. But in the end, the interesting questions are completely open, so it’s not as if you know coming into a story what the right answers are. I don’t think anyone should be scared away from the field by a lack of legal or technical background.
ROSENZWEIG: What’s the most fun part of writing in this area?
ROSEN: Learning about topics you don’t know anything about. It’s just an endlessly engaging subject. And each of these topics are ones that I hadn’t known much about, and realizing the complexities and the narrative quirks and putting it all together, and coming out knowing much more than you did before, and trying to share that with other people is very exciting. So, I think the fact that you don’t know what you think in advance but have to follow a lot of moving pieces of the puzzle is interesting. You just always feel engaged at having a chance to jump into these subjects and learn about them.
ROSENZWEIG: Best advice you could give a journalist who aspires to work in this area. You’ve mentioned patience and working the narrative. Any others?
ROSEN: I think of the journalistic advice that I always appreciated. Mike Kinsley’s sell every piece three times. That’s always a good idea, no matter what you’re writing about. I’d say it’s a good area to – it’s a growth area because it’s perceived as specialized. There are lots of wide-open opportunities. As long as people are not intimidated by the technical aspect, because they shouldn’t be, and realize that the basic principles can be grasped with some homework, it’s a wide-open field, and it would be great to have more people writing in it.
ROSENZWEIG: Mistakes. Where have you gone wrong that other people should, in looking into this, that you shouldn’t have – that in retrospect you wish you could have a do-over?
ROSEN: Some of my predictions have been wrong, as my predictions always are. Predictions are a hazardous business. In the “Naked Crowd” book, I think that I ended up arguing that although it was possible to imagine a thoughtful policy solution to many of the surveillance technologies I was talking about, from data mining to body scanners, the good balance would not be struck by the courts or the technologists on their own, but would have to come from Congress. And it’s now clear to me, 10 years later, that that was really simplistic, and that Congress is no more likely to adopt good solutions on its own than the technologists, the courts, or the regulatory agencies. It now seems to be a complicated interaction of all the various political and technological and judicial actors. As the case of the body scanners shows, you really need pressure from all of those things. You need political activism, you need a little bit of legislative muscle flexing, and ultimately the threat of the judicial solution as well in order to pressure the executive to adopt a sensible solution. So, I’d be wary of my own predictions and also say that each of these debates has evolved in such surprising ways. People get galvanized by unexpected things, and projects blow up in the government’s face in ways that wouldn’t have been anticipated that I just would be much more, have a lot more humility about predictions in general and think that the description and analysis of what’s going on is more useful than trying to be a pundit in this area.
ROSENZWEIG: That’s a useful thing. Anything else you want to say on this before we close it out?
ROSEN: I think that the area’s most interesting when it’s looked at holistically. So, government data mining is very important and raises interesting issues, but it shouldn’t be separated from the distinct but related issues raised by consumer data mining. The technologies are similar, the legal issues can be different, the values at stake can be different, but looking at the broad scope of how technology is affecting privacy across the range of public and private actors is really important. And there’s also this fascinating shift which suggests that today, Google and Facebook have more power over the future of privacy and free speech than the government, and the decisions that private corporations like Google and Facebook are making will impact the privacy debate more than the decisions of kings and presidents and Supreme Court justices. So, just looking at the government’s actions in isolation again would be too narrow, and it’s useful to look at the role of private corporations as well.
