Policy —

Geeks are from Mars, wonks are from Venus

A prominent computer scientist tells us what he learned from two years in DC.

Geeks are from Mars, wonks are from Venus

Two years ago, Ars described the appointment of Princeton computer science professor Ed Felten to be the Federal Trade Commission's first chief technologist as "a decision so shockingly sane that it's still a bit hard to believe." After 20 months on the job, Felten has wrapped up his tenure at the agency and returned to Princeton. He has been succeeded by Columbia computer scientist Steve Bellovin.

So what's it like to be a geek in the land of lawyers? Ars Technica interviewed Felten by phone on Tuesday to find out. First, an important disclosure: Felten was my advisor when I was in grad school at Princeton from 2008 to 2011.

Mr. Felten goes to Washington

Felten first came to public attention in the late 1990s when he served as the government's expert witness in the Microsoft antitrust case. During the aughts, he developed a reputation as a thorn in the side of companies that made insecure products, as he and his grad students found flaws in multiple DRM schemes. In one instance, he was forced to sue the recording industry to establish that publishing his research did not violate the Digital Millennium Copyright Act. More recently, he focused on exposing security flaws in touch-screen voting machines, announcing a proof-of-concept voting machine virus in 2006. We named him as one of our "people to watch" in tech policy in 2009.

Felten took a leave of absence from Princeton in January 2011 to serve as the first chief technologist at the FTC. He advised the FTC chairman about technology issues, assisted internal staff with technical issues that came up during their investigations, and served as a point of contact to the technology community. The agency had plenty of work for him to do, since it has responsibility for issues like consumer privacy and antitrust enforcement that frequently involve IT companies.

Computer scientists are a rare breed in lawyer-dominated Washington, DC, and Felten said it was sometimes a challenge helping policymakers understand the nature and limits of technology.

For example, he said a lot of people in Washington have a misconception that any problem "can obviously be solved if you try hard enough." Felten cited the SOPA fight (which the FTC wasn't directly involved in) as an example. "When the tech community says 'this won't work for technical reasons,' people hear 'we're not willing to solve it,'" he said. "A popular view of science is that it's going to eventually solve all problems, that no problem is unsolvable if you work hard enough at it."

Of course, an important part of computer science is understanding that certain things are impossible. "The idea that something is impossible or that it's practically impossible due to issues of compatibility and so on tends not to sink in" for many people, he said. They think that "any system can be broken, any system can be made secure, and it's just a matter of throwing resources at problems."

But regulators can also underestimate the degree of uncertainty about new technologies, he said. "There's this image of scientists and engineers as being super slow and meticulous about everything," he told us. They can be surprised to learn that "a product might have a behavior that's unexpected even to its developer."

Felten found that he could usually dispel these misconceptions once he got to know an individual policymaker. Especially within the FTC, once he "developed a relationship where they trusted me," he had an easier time persuading them.

The importance of relationships

Indeed, he said that the importance of relationships to policymaking is one of the key things about Washington that's not widely understood in the technology world. To effectively influence public policy, it's essential to know who the key decision-makers are, and what their idiosyncratic concerns are.

"It matters if a particular member of Congress or a particular policy maker has a certain background or a certain kind of experience or has a friend, a spouse, or a sibling who happens to work in a particular setting," he told us. "These things make a difference both in the level of understanding and attention that certain issues get, and also the ability to be flexible on certain kinds of issues."

And a lot of decision-making in Washington happens in informal settings. "People in the tech world often don't pay enough attention to things that are outside the public record," he said. A "huge amount of conversation" happens outside the official policy-making process.

"The backroom deal is seen as a sinister thing, but there are a lot of productive and positive conversations that happen," he said. These informal conversations can ensure that "people don't get surprised, things don't get misinterpreted. A lot of the process of policymaking in some areas is really about diplomacy in terms of interested parties—agencies, companies, industry groups, public interest groups"—hashing things out.

Still, that does mean that interest groups with the resources to hire full-time lobbyists have an inherent advantage. But Felten said there are ways ordinary geeks can influence the policy process. The most important thing they can do, he said, is to develop relationships with people who do have direct connections to the policy process. That might be your member of Congress, but it could also be people who work on public policy at your employer, state or local officials, or officers at membership organizations like ACM.

"Policymakers have all kinds of people telling them things all the time," he said. "Unless they have an understanding of how reliable your advice can be, they don't know how much weight to put on it."

Developing that kind of relationship takes time. And such relationships must be a "two-way street," Felten told us. For example, "You might want to talk to a policymaker, but there will also be times when they want to talk to you. Part of that relationship is being available and helping them when they want your help."

In the long run, Felten hopes to see more technologists have a formal role in the policy process. "We've had some degree of success in intervening when something big is up, but that's not a sustainable model," he said. He'd like to see technologists become "a community that is known and engaged consistently over time."

"I've thought about how other disciplines have managed to integrate themselves into the process," he said. "Economists are a good example. They've been very effective in convincing policymakers that their expertise is important."

It would obviously be foolish for senior government officials to make economic policy without seeking input from trained economists. Perhaps someday, it will be seen as equally foolish to make technology policy without a computer scientist in the room.

Channel Ars Technica