misleading unless we were able to effectively measure the relative importance of each Web
site. For example, the blocking of BBC or Wikipedia represents far more than the blocking of a less prominent Web site. Similarly, blocking a social networking site or a blogging
server could have a profound impact on the formation of online communities and on the
publication of user-generated content. While Internet users will eventually provide alternatives to recreate these communities on other sites hosted on servers that are not blocked,
the transition of a wide community is unlikely given the time, effort, and coordination required
to reconstitute a community in another location. At the other extreme, the blocking of one
pornographic site will have a minor impact on Internet life if access to thousands of similar
sites remains unimpeded. For these reasons, we have decided to summarize the results of
testing categorically, considering both the scope and depth of the quantitative testing results,
in conjunction with expert opinion regarding the significance of the blocking of individual Web
sites.
It is tempting to aggregate the results by summing up the scores in each category. Yet doing so would imply that the blocking of political opposition is equivalent to filtering that supports conservative social values or the fear of national security risks. These competing sets
of values suggest that a number of different weighting schemes might be appropriate. In any
case, the results are generally quite clear, as the most pervasive filtering regimes tend to filter
across all categories.
Country-specific and Global Filtering
A comparison between the blocking of country-specific sites and the blocking of internationally relevant Web sites provides another view of global filtering. Not surprisingly, we found that
Box 1.2
Where we tested
The decision where to test was a simple pragmatic one—where were we able to safely
test and where did we have the most to learn? Two countries did not make the list this
year because of security concerns: North Korea and Cuba. Learning more about the
filtering practices in these countries is certainly of great interest to us. However, we
were not confident that we could adequately mitigate the risks to those who would collaborate with us in these countries.
A number of other countries in Europe and North America that are known to engage
in filtering to varying degrees were not tested this year. This decision again was a fairly
easy practical choice. The filtering practices in these countries are better understood
than in other parts of the world and we therefore had less to contribute here. Many of
the countries in Europe focus their Internet filtering activity on child pornography. This is
not a topic that we will test for ethical and legal reasons.
20
Robert Faris and Nart Villeneuve
site. For example, the blocking of BBC or Wikipedia represents far more than the blocking of a less prominent Web site. Similarly, blocking a social networking site or a blogging
server could have a profound impact on the formation of online communities and on the
publication of user-generated content. While Internet users will eventually provide alternatives to recreate these communities on other sites hosted on servers that are not blocked,
the transition of a wide community is unlikely given the time, effort, and coordination required
to reconstitute a community in another location. At the other extreme, the blocking of one
pornographic site will have a minor impact on Internet life if access to thousands of similar
sites remains unimpeded. For these reasons, we have decided to summarize the results of
testing categorically, considering both the scope and depth of the quantitative testing results,
in conjunction with expert opinion regarding the significance of the blocking of individual Web
sites.
It is tempting to aggregate the results by summing up the scores in each category. Yet doing so would imply that the blocking of political opposition is equivalent to filtering that supports conservative social values or the fear of national security risks. These competing sets
of values suggest that a number of different weighting schemes might be appropriate. In any
case, the results are generally quite clear, as the most pervasive filtering regimes tend to filter
across all categories.
Country-specific and Global Filtering
A comparison between the blocking of country-specific sites and the blocking of internationally relevant Web sites provides another view of global filtering. Not surprisingly, we found that
Box 1.2
Where we tested
The decision where to test was a simple pragmatic one—where were we able to safely
test and where did we have the most to learn? Two countries did not make the list this
year because of security concerns: North Korea and Cuba. Learning more about the
filtering practices in these countries is certainly of great interest to us. However, we
were not confident that we could adequately mitigate the risks to those who would collaborate with us in these countries.
A number of other countries in Europe and North America that are known to engage
in filtering to varying degrees were not tested this year. This decision again was a fairly
easy practical choice. The filtering practices in these countries are better understood
than in other parts of the world and we therefore had less to contribute here. Many of
the countries in Europe focus their Internet filtering activity on child pornography. This is
not a topic that we will test for ethical and legal reasons.
20
Robert Faris and Nart Villeneuve