Brief 32: Anti-Americanism in the Middle East
EGAP Researcher: Daniel Corstange
Geographical Region: Middle East
Research Question:
Is Anti-Americanism in the Middle East policy-based or does it extend more generally to encompass all things American? Do people refuse to be interviewed depending on who they think is asking the question – governmental, academic, American, or non-American sponsors?
Preparer: Alicia Cooperman
Background:
Scholars are divided over the nature of Anti-Americanism in the Middle East: some posit that people recognize the difference between what America is and what America does, while others argue that people are generally anti-American.
However, most evidence of Anti-Americanism uses surveys to study public opinion; it may be biased if anti-American individuals are self-selecting out of the survey pool, since many surveys are done by American universities, foundations, and news agencies. In addition, we care about behavior, and not just public opinion, since anti-American behaviors have more immediate consequences for national security.
Lebanon is a diverse society in the Middle East with wide variation in religious sectarian groups, which have differing political views and levels of hostility toward the United States. While 71% of Christians hold favorable views of the U.S., only 12% of Shiites do so. Therefore, it is an interesting case to study anti-American opinion within and across groups.
Research Design:
This study discerns the type and extent of anti-American biases by measuring behavioral responses: whether or not subjects refuse to take the survey depending on who they believe sponsored it. In a mass attitude survey in Lebanon, subjects are randomly exposed to different survey sponsors, and their refusal rates are compared between treatment conditions.
Subjects were recruited with a script that began: “My name is [NAME] and I am from Information International, a survey research firm based in Beirut. We are conducting a short survey in the Greater Beirut area about current affairs in Lebanon on behalf of X.” The full recruitment script included the sponsorship manipulation twice, once at the beginning of the script and once at the end, with the following treatment conditions:
- American Embassy
- Canadian Embassy
- the University of Maryland (a major research university in the United States)
- the American University of Beirut (a major research university in Lebanon)
- the American University of Beirut and the University of Maryland (which are major research universities in Lebanon and the United States)
- (Control: no sponsorship mentioned)
In Lebanon, the University of Maryland is a generic university that does not have any name recognition. While the American University of Beirut has “American” in its name, it is one of the most religiously and politically diverse universities with a well-known reputation as a premier institution in the Middle East. In the local context, it does not imply any American affiliation and is politically neutral.
Through face-to-face interviews in Beirut, Lebanon during June and July 2011, researchers approached 2,592 households without mentioning sponsorship, of which 96% (2,481) permitted the survey. Of the households who permitted the survey, 85% (2,100 of 2,481) continued with the survey after hearing the recruitment script with the sponsorship descriptions. After the survey (or refusal), the respondents were debriefed verbally and in writing to correct the use of deception.
Results:
The findings are quite definitive: university sponsorship did not change participation relative to the control group, while both embassy conditions induced strong refusals. Subjects acted very differently toward an American university vs. the American government. In addition, subjects were not solely anti-American; they also were anti-Canadian, though to a lesser degree. The control group and academic groups all had approximately 10% refusals, while the Canadian embassy condition had 21% refusals and the U.S. embassy had 33% refusals. Anti-American trends were similar across sectarian groups with different base levels of hostility and political preferences. In absolute terms, subjects in Shia (anti-American policy) neighborhoods had higher refusal rates than in non-Shia neighborhoods; however, subjects in Shia and non-Shia neighborhoods were similar in their ratios of rejecting surveys sponsored by the American Embassy vs. University of Maryland or American Embassy vs. Canadian Embassy. Specifically, within Shia neighborhoods, the academic and control treatments had equivalent refusals (11-14%) while the embassy treatments had very high refusals (42% for Canadian and 57% for American). The non-Shia group was similar in having lower control/academic refusals (4-8%) than embassy refusals (13% for Canadian and 22% for American).
Policy Implications:
This work contradicts the many bombastic statements in the media claiming that there is generalized anti-Americanism sentiment in the Middle East. The results suggest that responses to anti-American views are based on policy and not a generalized view of America.
Individuals across sectarian and political views were able to differentiate between an American university and the American government, and they differentiated slightly between Canadian and American governments. All subjects participated equally in surveys sponsored by American or Lebanese universities.
The equivalent results, even in Shia neighborhoods that espouse anti-American preferences, suggest that even “radical” groups are able to differentiate between American policy and America more broadly. Still, additional survey experiments are needed to validate these findings outside a diverse, capital city and outside Lebanon. Other treatments such as non-Western government and Lebanese government sponsors or different topics would also prove interesting.
These findings are also good news for researchers doing survey work in the Middle East. Potential concerns of selection bias, where anti-American subjects might refuse to participate in surveys by American universities or non-governmental organizations, is unlikely.