Twitter and the Arab Spring: New Evidence

John and I are members (along with Sean Aday, Deen Freelon and Marc Lynch) of a multi-author team which has been doing USIP sponsored work on new technology and the Arab Spring. The USIP has just published a report based on our findings, which suggest that Twitter was rather more important as a way of conveying news out of the region and towards an international audience, than as a means of communication within or across the affected countries.

If the audience is primarily local, then information consumption is more likely to play a significant role in mobilizing protest or affecting dynamics on the ground. If the audience is regional or even global, then new media is arguably functioning as a megaphone, generating external attention from citizens, news media, and governments outside of the country. The mechanism by which it achieves this function might be called bridging. Bridging is particularly valuable when traditional news sources are unavailable or regulated by the government, or when violence or repression makes it difficult for mainstream media to directly cover an unfolding story. For example, the Tunisian government tightly controlled themainstream Tunisian media, making information available online all the more valuable. But in more open polities, which tolerate a higher degree of press freedom within their authoritarian
structure, the Internet would become only one part of a complex media ecosystem. New media and user-generated content may also have mattered disproportionately in relatively distant cases such as Yemen, where few mainstream media had bureaus, or Syria, where first extreme repression and then extreme violence made it difficult for foreign journalists to operate.
The location-specific patterns in the data examined thus far—temporary spikes in clicks that correspond to widely publicized events—suggest that traffic largely consists of consumers outside of the countries where the protests took place. That is, it would appear that, in part via Twitter, was functioning like a megaphone. For each protest event, we divided those who clicked on links into three categories—those located within the country, those located outside of the country but still in the MENA region, and those located outside of the MENA region—and calculated the proportions of clicks from each category (figure 2).
During all four of the protest events, the majority—and at times the vast majority—of clicks came from outside the country. Only in the case of Bahrain did a large percentage of clicks (37 percent) come from within the region. This is not a surprise, given that Internet penetration is much greater in many countries outside of the region than within the region. But the lack of surprise does not make the pattern any less important.

8 Responses to Twitter and the Arab Spring: New Evidence

  1. Olaf July 11, 2012 at 4:37 pm #

    It would have been interesting to see a comparison between the clicks in absolute figures. The internet penetration is important of course, but couldn’t it perhaps be said that just in case the same group of “outsider” people are the most frequent clickers in all the cases then the absolute number of clickers in each of the countries should be the figure that should be compared?

  2. Prison Rodeo July 11, 2012 at 5:58 pm #

  3. Kevin P July 11, 2012 at 9:00 pm #

    from the COIN principle: ‘whenever possible, the one must fight on the ground where one is stronger and one’s opponent is weaker’ … the people found voice and now we can see (or hear) them in their sanctuary.

  4. QS July 12, 2012 at 4:02 am #

    This doesn’t really tell us much, aside that there are more internet users in the non-MENA region than within it, which is of course quite obvious. To me, it says that the world was paying attention via the Internet. But that isn’t surprising either.

    What would be of greater interest is the “quality” of intra-country twitter usage, e.g. the magnitude of utility and effect of online messaging, rather than the pure quantity of clicks. The “clicks” of some guy at the internet cafe in Aswan regularly following the feeds but not engaged in the revolution are of much lesser importance than the “clicks” of those who actively make use of the information, say Cairenes engaged in mobilization of persons.

  5. Monkey C July 12, 2012 at 9:53 am #
    oh the irony
    TLD i sLibya

  6. Michael Bernstein July 12, 2012 at 12:31 pm #

    How does this look if you normalize the clickthrough numbers by the internet-using population in each country? In absolute numbers, these countries have much smaller populations than the rest of the world, so it’s not surprising they might be overwhelmed…

  7. Kevin P July 12, 2012 at 2:43 pm #

    @QS…. “This doesn’t really tell us much, aside that there are more internet users in the non-MENA region than within it, which is of course quite obvious.”

    It tells us they are paying attention. Voice is an indicator, same as ennui

  8. Zeynep July 13, 2012 at 8:28 am #

    There is a lot of good in this report but let me clarify a few things. I’m seeing this (mostly elsewhere) represented as a measure of use of new media–and that it was mostly used by outsiders. But that is not what the data measures. It also does not even measure twitter use per se. It’s also not a measure of the whole new media ecosystem–which, as the report notes is totally integrated into media ecology (I’ve written about this in a Journal of Communication article which looked at media use among Tahrir protestors: Email me if it is paywalled for you ).

    Rather, this report is a good measure of one sub-component of a sub-component of this system: It’s a measure of on Twitter on tweets with certain hashtags. Thus, it doesn’t even have Tweet data without links shortened using a particular platform or tweets without particular and specific hashtags. (I do not mean to belittle the data at all! I think it’s great to have it and congrats to my friends and colleagues for this report; I’m adding context to its scope). I wish we had more of this kind of analyses–especially for Facebook and Youtube which remain underanalyzed due to data access issues. So we all go look at Twitter because it is accessible (I do too!) but we have to avoid the looking for keys under the light fallacy. (Again, I’m talking about the way this report is being presented which is not really the fault of the authors though I would have phrased the executive summary differently to make some contextual variables clearer),

    My observation do concur with some of the key findings so I am not disputing the specific results as long as they are understood in correct context. I agree –and have long said– that Twitter was used mainly as a bridge to outsiders–but not solely. It is also used for protest coordination (yes, really, you only needed to be on Twitter during the #maspero clashes to see this) and for broadcast, and banter. It totally makes sense for the tweets with shortened links to be accessed by outsiders.

    Plus, I think frequency analyses like this one (how often was X used) are very important as long as one understands that this is only one measure of importance and place in the ecosystem. There are times when a particular accordance is rarely used, but makes all the difference if it is. Similar with Internet penetration analyses. Yes, it matters what percent of a country uses Internet. But it doesn’t make Internet unimportant if penetration is, say, 20 percent. It just means that it fits into a particular role among particular groups; it does not say it has no role. (Most revolutions have scant participation; historical evidence seems to be that it is usually 1-2% of the population on the streets with its height perhaps in Iran 1979 with about 10%).

    Finally, I think being able to break censorship and command your own narrative to the outside world seems like a pretty big deal to me. I know many revolutionaries in previous uprisings elsewhere would have been very happy to have this affordance.

    Then there is the role of Facebook in mobilizing and making dissent visible to social networks, the role of Youtube in exposing corrution and torture, the use of blogs and the myriads of forums that have popped up in the region, the undernoted use of IRC which many key activists still do use and have been using since the nineties, the way new media feeds into the broadcast/big media ecology (very integrated)…