Attitude problem: what social media can’t tell us about Palestine

The last time you visited your favorite blog, how wide of a cross-section of public opinion did the comments represent? It probably depended on the blogger, on the article, and on the mood of the day.

Yet these limitations haven’t stopped advocates from trying to discern Palestinian public opinion from bloggers’ views. Last week, the Foundation for the Defense of Democracies (FDD) presented to Congress a new report that puts Palestinian public attitudes, in contrast to polling data, in a decidedly hostile light. "P@lestinian Pulse: What Policymakers Can Learn From Palestinian Social Media," by Jonathan Schanzer and Mark Dubowitz is the first published study which attempts to ascertain Palestinian public opinion exclusively from web sources. But is the report accurate?

FDD contracted out to ConStrat, a D.C.-based communications firm, which mined an array of content, and then FDD drew broad conclusions such as Hamas "supporters showed no apparent disagreement with Salafists such as al-Qaeda" and "Palestinian reform factions are weak and have little influence online." But the study’s methodology leaves much to be desired: it’s impossible to confirm whether the sample only includes Palestinians; there isn’t a clear theory of how to analyze this content or how FDD reached these conclusions.

Sentiment analysis via social media is a rising trend in the field of public opinion. In the Arabic-language context, Harvard’s Berkman Center for Internet and Society published an extensive report that portrays the Arabic blogosphere as a dynamic, complex ecosystem. One of the Berkman Center’s key findings was that they "found little support for terrorism or violent jihad in the Arabic language blogosphere and quite a lot of criticism." This doesn’t mesh with Schanzer and Dubowitz’s conclusion that participants in Arabic-language social media represent a growing radicalized population that seeks to spoil Obama’s peace push.

Online social forums should tell Palestine watchers a lot. But they won’t necessarily offer a more nuanced picture of the 4.7 million Palestinian refugees living in the Occupied Territory and also in Lebanon, Jordan, and Syria who are less likely to have internet access than those in the West Bank. Moreover, the lines between those residing in Palestine and those in the diaspora are blurred online.

Insofar as Arabic-speakers across the region are passionate about Palestinian issues, it’s difficult to extract a social media environment which is distinctly Palestinian. In fact, the Berkman Center wrote about the Levantine/English bridge, which includes "Lebanese, Jordanians, and Palestinians… connect[ed] to the United States and international blogosphere by ‘bridge bloggers’… There is no hard division in the network between English and Arabic blogs."

The authors of "P@lestinian Pulse" rely on ConStrat’s "military grade technology," which retrieved 1,788 statements from 1,114 unique posts situated in 996 online threads. A single Arabic-speaking analyst from ConStrat reviewed that entire basket of content. The authors then organized this information into six pre-selected themes or "nodes." The nodes — "peace process, violence, theological radicalization, political radicalization, Palestinian reform factions, and outside radical influences" — were apparently chosen with an end result in mind. Interestingly, topics such as "Israeli occupation, Israeli violence, settlements, and resistance," which presumably make up much of Palestinian concerns on any given day, were not included in the study’s taxonomy.

But how can we even verify that the participants whose posts were analyzed in "Pulse" were Palestinian? The authors write that adequate statistics do not exist to establish who is exactly participating in online Arabic-language discussion forums, though it’s assumed that participants are "generally educated." Nicole Nix, director of cyber research at ConStrat who served as the lead consultant for FDD’s study, said, "We didn’t provide numbers because that wasn’t really required, but we gave our impressions of how many people actually were Palestinian." Unlike other web-centric studies, "Pulse" neither offers percentages for the opinion trends identified nor provides appendices containing the full set of content reviewed.

In assessing their findings the authors use the phrase "on the ground" eleven times to describe circumstances in Palestine. Yet, according to both Nix and Schanzer, no Palestinians were formally consulted during the report’s production.

Dismissing the existing cannon of Palestinian polling data, Schanzer and Dubowitz assert the lack of support for third-party leaders in the blogosphere. This contrasts with the Palestinian Center for Public Opinion‘s August survey in which a 47 percent plurality preferred "a government with a majority of independents" over a government with the majority of Hamas or Fatah. Surveys suggest that Palestinian players have great resonance on the ground in spite of their absence online, or in spite of the author’s inability to find them.

Nix offered a reason for such disparities: "For this particular study we didn’t track it against polling data, we didn’t track it against focus groups, or anything like that."

With all of these methodological challenges, it’s not at all clear what U.S. policymakers can learn from Palestinian social media. One of FDD’s policy proposals is that the White House should increase funding for the Department of State’s Digital Outreach Team (DOT), which engages with Arabic-language social media users. Yet Schanzer and Dubowitz also write, based on their nine weeks of monitoring sites where DOT plays an active role, "the State Department’s efforts to influence online discussions were largely ineffective." State’s Bureau of International Information, in which the DOT is situated, did not respond to phone calls.

"We don’t understand the extent to which Palestinian social media is popular or is representative necessarily of the wider Palestinian population," Schanzer, a former counterterrorism analyst at the Department of the Treasury, told me by phone. "What we do understand is that somewhere between 4-20 percent of the Palestinian people use the internet, the percentage of those who actually engage in discussions in online media could be much smaller."

It’s quite a tautology. These selected online threads shed light on a single cluster of Arabic-speakers: those who air their laundry on the internet.

Khalil Shikaki, director of the Palestinian Center for Policy and Survey Research, was skeptical as to whether one could approach a study of Palestinian blogs in a scientific manner: "The idea of having a representative sample by looking at the internet is absolutely ridiculous."

To be sure, Palestinians are questioning of the peace process as it is and are not confident that a Palestinian state will be established in the near term. Yet a majority of Palestinians prefer a two-state outcome with a state of Palestine alongside Israel. By focusing solely on "rejectionist" posts from so-called "Palestinian" social media, the political landscape is distorted.

Still, the exponential rise in access to technology in the West Bank and Gaza, compounded by the difficulty of movement between Palestinian locales, highlights the importance of social media and the need for further study.

Adel Iskander, a global communications expert at Georgetown University’s Center for Contemporary Arab Studies suggested that the expansion of internet usage in the Palestinian territories is directly correlated with their desire to communicate with the world. "The only way to understand the Palestinian online presence is to look at in anthropological terms — what is it that they are doing online and what are they looking for — rather than how we can define policy around them."

Indeed, Fatah and Hamas activists are competing for attention and attempting to build communities in cyberspace. But for Schanzer and Dubowitz to say "ConStrat analyzed the Palestinian social media environment," is in itself a leap of judgment: there are, of course, numerous communities and no monolithic "Palestinian social media environment." Beyond that, to assume that online conversation reflects public attitudes only contributes to the misreading of Palestine’s "pulse."

Jonathan Guyer is an assistant editor for The Middle East Channel and blogs at Mideast by Midwest.