Harvard University’s Berkman Center has come out with an interesting report
on the usage of circumvention tools, which allow users to access Internet content that’s otherwise unavailable to them.
The most important finding is that “overall usage of circumvention tools is still very small in proportion to the number of Internet users in countries with substantial national Internet filtering.” According to the authors' estimates, the number is no more than 3 percent. (The report gives a really good breakdown/primer of the various circumvention tools out there.)
There are a number of problems that prevent people from using proxies. One is that they don’t work very well, as governments block them or their performance speeds are too slow. Instead of cat vs. mouse, it’s more like whack-a-mole, where governments block proxies only for them to reappear with slightly altered domain names.
Then there’s the problem of technical expertise. Some of the simpler proxies work by a user just going to a website and entering the domain address they want to visit into a box. Others, however, require fiddling around with network settings, something many users can’t or won’t feel comfortable doing.
Increasingly, there’s the question of trust. Who exactly is running that proxy? A criminal enterprise that doesn’t care about you accessing foreign news, but just wants to steal your banking information? Or even worse for some, the proxy could be run by the very government whose censorship you are trying to evade. As the report notes:
Even though it is unlikely that a Chinese government agent or another filtering government is running most of these HTTP/SOCKS proxies, it would be surprising if a Chinese government agent were not running at least some of them given the extreme ease of setting one up.
The policy question, in particularly for the United States which supports circumvention tools in various capacities
, is whether it's worth putting money into circumvention tools when so few people use them. The Haystack affair
has left many with skepticism about circumvention.
As Ethan Zuckerman has pointed out before in "Beyond Circumvention"
, one of the problems with proxies is that "spammers, fraudsters, and other Internet criminals use proxy servers to conduct their activities, both to protect their identities and to avoid systems on free webmail providers, for instance, which prevent users from signing up for dozens of accounts by limiting an IP address to a certain number of signups in a limited time period." So for Western states, there could be some tricky funding questions.
There are other complications for well-meaning governments. In the same way that some NGOs working under repressive regimes don’t want the “taint” of U.S. government funding, the report suggests that official governmental support of circumvention tools -- and the increased press coverage they receive -- could give them the kiss of death and increase the chances of being filtered.
Ironically, it might be the dodgiest proxies (in terms of the shady characters they're run by) that are the most successful at evading the filters. The authors said they were surprised to find that many widely used proxies in “highly censorious” nations remained unblocked for long periods of time. As the report notes, these tools have not been “lauded much if at all in the U.S. press as agents of political change.” So basically a rag-tag army of proxy providers, supporting themselves with Viagra ads, might be more effective than a big government-funded organization. Food for thought.
One caveat though, and from someone working in U.S. international broadcasting: Three percent sounds like a small amount, but is it really?
The report finishes on the somber note that users may not be particularly interested in circumvention tools because they’re quite happy with what they’ve got: i.e. despite the filtering, they still prefer content in their own language about their own local interests.
For the vast majority of users, largely apolitical who use the Internet to watch movies, hang out with friends, send email etc., that's likely true. But there’s also the significant segment of actively politicized people (normally a pretty small percentile) that, for instance, U.S. international broadcasting is interested in targeting. The change agents and the multipliers are more likely to also be the proxy users. In policy terms, it's about quality not quantity.
The hope is that as circumvention tools become easier to find and use in local languages and are more accessible for non-techies, the numbers of people using them will grow. The catch, of course, is that they might be used more by people wanting to watch Hulu in Mexico than by people living in closed societies who are hungry for information.