Is Iraq Actually Falling Apart? What Social Science Surveys Show

By Mansoor Moaddel

Several nationally representative surveys carried out in Iraq between 2004 and 2013 provide important facts about Iraqi orientations toward secular politics, basis of identity, Americans, and Iranians. These facts have serious implications for the territorial integrity of Iraq, support for an Islamic government, and the U.S. policy toward the country. These surveys have shown evidence of:

(1) Support for Secular Politics: A much higher percentage of the Sunnis, even higher than the Kurds in some years, believe that Iraq would be a better place if religion and politics were separated. This support has increased from 60% in 2004 to more than 81% in 2013. By contrast, support for secular politics among the Shia has an inverted U-shape between 2004 and 2013. It went up from 44% in 2004 to 63% in 2011, and then dropped to 34% in 2013.

From the standpoint of public opinion, this evidence implies that the cooperation between the Sunni tribes/groups with ISIL (Islamic State of Iraq and the Levant) may not indicate mass conversion to religious extremism. Rather, it is driven by a common hatred of the Shia sectarian government ruling the country.

(2) Recognition of Iraq (and not religion) as the basis for identity: The Sunnis and Shia converge in defining selves as Iraqi, rather than Muslim or Arab, above all. This support rose from 22% in 2004 to 80% in 2008, and then dropped to 60% among the Sunnis. Among the Shia, it was 28% in 2004, increased to 72% in 2007, and then dropped to 62% in 2013. There is not much support for Iraqi identity among the Kurds. Among the Kurds, on the other hand, there has been a shift from predominantly Kurdish identity to religion.

Reinforcing attachment to the nation rather than to the religion of Islam in politics is the fact that both the Sunnis and Shia (1) prefer politicians who are committed to the national interests over politicians who have strong religious convictions by at least a factor of 4 to 1, and (2) consider a good government one that makes laws according to the wishes of the people over the one that implements only the sharia by at least a factor of 3 to 1.

These findings indicate that neither the Shia nor the Sunnis would be interested in the partition of Iraq or in the implementation of a religious state, and the current sectarian struggle is for political supremacy rather than division. Iraq may converge to the Lebanese model of sectarian strife.

(3) American Unpopularity: Well over 90% of the Sunnis and 86% of the Shia do not wish to have Americans as neighbors, with these figures remaining largely stable over the last decade.

(4) Iranian Unpopularity: Iranians have become quite unpopular among the Sunnis. Fully 88% of the Sunnis in 2011 did not wish to have Iranians as neighbors. Among the Shia, Iranians were not as popular as might be expected, however. Those Shia who did not wish to have Iranians as neighbors fluctuated between 44% and 73% between 2004 and 2011.

These figures have implications for the U.S. efforts to combat terrorism in Iraq and assist Iraqis in maintaining territorial integrity of their country. There is certainly a unity of purpose between U.S. policy makers and the Iraq public. Neither side wants an Islamic government nor wishes the partition of the country. The U.S., however, is facing two ruthless enemies with proven capacity to murder Americans; ISIL and other Sunni terrorist groups, on one extreme, and the Quds Force of the Islamic Republic’s Revolutionary Guards, on the other. Given the U.S. unpopularity, evidenced in part by the consistently negative opinion of Americans as neighbors, whatever the U.S. does is suspect from either the Sunni or Shia standpoint.

The US must pursue a three-pronged strategy which: (1) weakens the ties between ISIL and the more moderate Sunni groups, (2) weakens the ties between Iraqi Shia and the sectarian Islamic regime in Iran, and (3) strengthens ties between moderate Sunnis and Shia within Iraq.


Between the 2004 and 2013 surveys, the Sunnis expressed much more support for secular politics than the Shia or, recently, even the Kurds. The percentage of the Sunnis who strongly agreed or agreed that Iraq would be a better place if religion and politics were separated was 60% in 2004, 86% in April 2006, 80% in October 2006, 74% in 2008, 82% in 2011, and 81% in 2013. These values for the Kurds were similar to the Sunnis; 69%, 86%, 68%, 77%, 63%, and 75%, respectively. Among the Shia, however, support for secular politics were much lower, having an inverted U-shape; 44%, 42%, 60%, 63%, 62%, and 34%, respectively.



Since the 2004 survey, the percentage of the Sunnis who defined themselves as Iraqis, rather than Muslims or Arabs, above all, has increased considerably. This percentage was 22% in 2004, 24% in 2006, then 53% in March 2007, 58% in July 2007, 80% in 2008, 68% in 2011, and 60% in 2013. Among the Shia, support for Iraqi identity has also increased since 2004; it increased from 28% to 32%, 62%, 72%, 67%, 64%, and 62%, respectively. Iraqi identity, however, is much lower among the Kurds; the percent Kurds who defined themselves as Iraqi above all fluctuated between a low point of 5% in 2006 and a high point of 20% in March 2007.



Iraqis were asked to choose between two definitions of a good government (1) one that implements only the law of the sharia, and (2) one that makes laws according to the people’s wishes. A great majority of the respondent opted for the second option. That is, 86% versus 14% among Sunni Arabs, 75% versus 25% among the Shias, and 91% versus 9% among the Kurds in the 2011 Iraqi survey demonstrated support for a secular rather than religious government.



Given the choice between the two statements that (1) it would be better for Iraq if more people with strong religious view held public office, and (2) it would be better for Iraq if people with strong commitment to national interests hold public office. Here again, national interest trumps religion, with 89% versus 11% among Sunni Arabs, 81% versus 19% among Shia Arabs, 60% versus 40% among Kurds in 2011 Iraqi survey showing support for secular politicians.



Americans have remained quite unpopuar in Iraq. Among the Sunnis, the percentage of those who did not wish to have Americans as neighbors fluctuated between 92% in 2011 and 99% in 2006; among Shia between 86% in 2011 and 98% in 2006; and among the Kurds between 46% in 2006 and 69% in 2011.



Iranians have also grown unpopular among the Sunnis, as the percentage who did not wish to have Iranians as neighbors grew from 72% in 2004, to 91% in April 2006, 97% in October 2006, and then dropped to 88% in 2011. These values for the Kurds were 53%, 46%, 84%, and 69%, respectively. Among the Shia Iranians were not as popular as might be expected. Those Shia who did not wish to have Iranians as neighbors fluctuated between 50% and 73% between 2004 and 2011.


Mansoor Moaddel is Professor of Sociology, University of Maryland

8 Responses

  1. There is clearly a mistranslation of the word “neighbor”. What was the original word in the survey in Arabic and what were its implications? We know that surveys, with careful wording, can me made to show most anything.

    • Nonsense. People who conduct survey research generally are clued in when their colleagues are cheating. I don’t remember what word is used for the Arabic. I think this is a World Values Survey question that gets at who you like based on whether you would find it acceptable for them to be next door. The literal Arabic word is Jar but I don’t recall whether the translation of this question uses that word or a phrase to get at that point. If I remember I’ll look it up when I can but no access here. But to your point – the term used is commonplace, benchmarked, understood correctly, etc. in short – not the problem. The problems are as I identify below.

      • من هذه المجموعة جيرانك is the Arabic phrase used often in this type of question. although I do not have the wording for this specific questionnaire it is no doubt from this family of questions. So yes it is likely using the Arabic word for neighbor, or your neighbors (jayranak).

  2. Cooperation may not indicate mass conversion to religious extremism of local Sunnis supporting the insurgency, but it does indicate easy condoning of known extremist terrorist groups that want to create a sectarian religious state, due to an affinity to the same shared sect background and religious-political goal. I worry its not restricted to just a common sectarian hatred for the Shia govt, but against the population as a whole.

    The false equivalence of comparing the takeover of Sunni Islamist terrorists to Al Quds drawn presence as equal ruthless enemies to the US, is like claiming that Democrats and Republicans are polar opposites but equal extreme war-mongers detrimental to the US.

    ‘Weakening ties’ is said like as if its the same reasonable applicable solution for ISIL, a terrorist group, and Iran, a religious state. It shouldn’t be a weakening of ties between the Moderate Iraqi Sunnis left and local/foreign extremist ISIL terrorists, who’ve escaped the sectarian label, but a ‘cutting’ of ties. Weakening ties would be applicable for Iraqi Sunnis with sectarian religious Saudi Arabia and the Gulf, who don’t get mention despite evidenced financing of ideological global and regional radicalism, and indeed more detrimental than ‘sectarian’ Iran in that regards. However, Iran does need to step back (after a solution for this crisis), and should be a promotion of Iraq’s Najaf or secularism as an alternative, but of course there’ll be ties by the very nature of the holy cities in Iraq and pilgrims.

  3. I am not sure how much to trust Face to Face surveys in areas like Iraq. It may be that areas with more hardline views are systematically underrepresented because interviewers would be redlined out of going there. Coverage error is so large a factor in conflict zones that it is hard to see the results as anything more than a “plausibility probe.”
    Is there an online source of the questionnaire and the methodology? How were households sampled? Was in-house respondent sampling done or only speak to head of household using a male team leader interviewer? Was statistical analysis done on the results to exclude cases that signal likely interviewer cheating (as far as it can be known – interviewers attempt to appear random but are never successful when they do this – there are techniques to uncover this)? The answers to these questions are important.

  4. Some of the findings that I find difficult to reconcile here are that while Kurds have typically shown broad support for secular politics and demonstrated a proportionally greater interest in the ‘will of the people’, they still showed quite strong support for government officials with a strong religious conviction. I wonder whether these individuals see a religious man as more morally inclined and therefore more likely to be an effective administrator etc. also raises the issue of their ability to truly separate the two in application.

Comments are closed.