Is Circumvention Really So Important?

Yesterday, the Berkman Center released a paper that attempts to estimate circumvention tool usage globally.  As one of the lead authors, Ethan Zuckerman, explained on his blog yesterday:

We were specifically interested in trying to compare usage of different types of tools – sophisticated blocking-resistant tools like Tor and Ultrasurf, ad-supported web proxies like Proxeasy or HideMyAss, and VPN-based systems like Hotspot Shield and Relakks. Unlike in our previous study of some of these tools, we weren’t trying to compare the functionality of these very different tools, or evaluate their performance – we just wanted to answer the question, “How many people use this tool?”

More broadly, and like Ethan, I’m interested in this question: “How many people are using circumvention technology at all?”  The answer we came up with, which you can certainly argue with, is about 3% of all Internet users.

Three percent of all Internet users.  Not very many.  What we don’t have the answer to is “why?” – Why so few users?  Is it that the vast majority are unaware of circumvention tools, or is it that they simply don’t matter that much to the majority of users?  We hear so much about Internet filtering, but could it be that the vast majority of users can find what they need without circumventing it?

As Ethan points out, that may be the case.  There are definitely people who require unfettered access (something that the Cuban government recognizes, incidentally), such as academics and activists, but how important is it for the average person to get to Facebook?

I’ve got another hypothesis that I’m hoping our upcoming survey of attitudes and perceptions toward Internet filtering and circumvention tech will help confirm: Most circumvention tools simply don’t work the way users wish they would.  I remember being in Morocco in 2005, at the beginning of my blogging career.  I was still using LiveJournal, and was surprised at my first login attempt to learn that it was blocked in the country (it still is).  So I googled “proxy” and picked the first one off the list.  I could get to just fine, but any attempts to login simply failed.  Most free proxies, at the time anyway, aren’t set up to keep users logged in to dynamic sites.

There are a host of other options of course, but short of paying for a VPN (something out of reach financially for many of the world’s netizens), they all have their flaws.  Tor’s too slow, some tools block videos because they take up too much bandwidth, still others block certain categories of site.

Of course, this doesn’t account for the vast majority of non-usage, but it’s yet another thing to think about.

I’m excited about this research – with all of the frenzy over filtering and circumvention, it’s all too easy to forget the human side of things.  And this is very much a human issue.  We’re talking not about tools and sites but about people and what they want or need to access online.

15 replies on “Is Circumvention Really So Important?”

It has to do with *Access* to knowledge and the way hyperlinked web works.

While proxies are available in most countries, you can’t say that information is accessible if you have to go through the hassle of activating a proxy to access it. You have to value a particular info a lot to make the effort. The problem is : if you can’t access it, how do you value it? So most of the time people, logically, don’t make the effort. Censorship is effective.

The second problem is, now the web accounts for most of the internet experience of everybody. People exchange links : when I stumble upon an interesting link, I re-share/re-publish it to my social network. When I stumble upon a censored link, I don’t share it. For example here in Tunisia people never share Youtube links (youtube is censored) so it’s like youtube does not exist. The consequence is Tunisians rarely stumble upon a Youtube links, given the composition of their social network.

Excellent point, Slim – though it’s possible to argue that those YouTube links don’t matter, it certainly takes the joy out of discovery on the web!

Personally, I think that more interesting than the fact that only 3% of internet users are using circumvention tools is what they are using them for. If it’s true, for example, that only 3% of Mexicans are using circumvention tools then I must know all of them … and I’d say that 99% of those users (including me) strictly use them to access geo-blocked sites and geo-blocked content within sites like YouTube.

Geo-blocking of content is probably the best thing that has happened to makers of circumvention tools.

A few points, you claim “Tor is too slow”, this was true 2 years ago, not today, see for results.

Tor will always be slower than single-hope proxies, because Tor’s goal is to protect you vs. just circumvent the local censorship.

3% of internet users is roughly 60 million people. (This assumes the ITU is correct in stating 2 billion people online or this While still a small percentage, 60 million people seems larger than expected. Tor alone has had 24 million downloads of the software in 2010.

Thanks Andrew –

I should have been clearer. As an occasional user of Tor, I don’t find it too slow. Nevertheless, that remains the number one complaint from my friends abroad, particularly in Arab countries with slower Internet speeds. That said, I should also have stated that many of them use it regardless, because of safety.

I’m not sure I understand your second point – If Tor had 24 million downloads, then why does 60 million proxy users seem larger than expected?

I find that interesting as well – last year, I had a roommate who was a public school teacher. She came home one day with a funny story – her students, tired of being blocked from Facebook, were using a Turkish IP address to get around the school’s filtering (particularly hilarious for those of us who know the Turkish filtering landscape). And Hulu! Nearly everyone I know outside the U.S. uses a VPN to watch their favorite U.S. shows.


I posted too early and then couldn’t go back and edit the comment.

To back up a statement, is it 3% of users in heavily filtered countries or 3% globally that use proxies? 60 million seems like large number of circumvention tool users (for all tools). Tor’s 24 million downloads does not equal individual users. My original, and lost, point was that at the scale of 24 million things (downloads or users) means we may have nearly half the circumvention users, which sounds very high. Maybe my expectations are wrong.

I’m curious about this survey you mention. Wouldn’t it be difficult to do a survey about attitudes about circumvention techniques? I assume you’d ideally like to survey internet users that are not using said techniques, which might make it hard to reach them in the first place.

Hi Kamal,

Sure – the survey is not yet public but will use a sample of approximately 4,000 people (bloggers, to be precise) in 18 different countries. The list of countries includes some which filter the Internet heavily, some which filter minimally, and a couple which don’t filter at all. Included in the sample will, presumably, be people who use circumvention techniques and those who don’t.

Undoubtedly, the sample will contain more “net-saavy” individuals than the general population but will hopefully still give us an idea of how people in different countries perceive and utilize circumvention tools.

As to your latter point, I’m not sure why you think people who don’t use circumvention would be harder to reach. Even people living in countries like China and Iran, where Internet filtering is pervasive, still have considerable access to the global Internet, thus we have considerable access to them.

Hi Jillian,

Thanks for the info. As for your query at the end, I have to admit I was thinking from my own viewpoint. Several U.S. based human rights organizations have websites that are blocked in countries like China. Depending on how they normally do surveys, that could make it a bit harder to get the surveys out to everyone intended (if those surveys normally reside withing their organizational websites). I suppose there are simple ways around it, though, like using a 3rd party survey service that doesn’t bear your domain name and thus doesn’t get filtered.


Good point, Kamal. I don’t want to share too much of this publicly, but happy to talk shop via e-mail if you’re interested in how we’re getting our survey out.

Censorship, either way is a BAD thing. Look at the more extreme cases, like China. I really think that in the next years VPN services and proxies will get more and more popular. I use for some time now, it`s working great, but I would just like to be able to access any site I want without any techie stuff…

Leave a Reply

Your email address will not be published.

This site uses Akismet to reduce spam. Learn how your comment data is processed.