Another 16% said they both reek. Righties will find cause for hope in those numbers, lefties will wonder why after four years we can’t get a clear majority to prefer life under American occupation to life under the Arab Stalin. Glass half-full, glass half-empty.

The whole poll is like that. You’re better off with the summary at the polling firm’s website than the Times of London article, which misleads a bit in reporting that only 27% of Iraqis say the country is in a state of civil war. That’s accurate as far as it goes, but another 22% say they’re “close” to a civil war but not there yet. TOL neglects to mention that.

If you can spare the time and concentration, it’s worth flipping through the crosstabs to compare the numbers by sect and ethnic group. Keep in mind that “Sunnis” encompasses both Arabs, most of whom are militantly opposed to the U.S. presence, and Kurds, most of whom are militantly in favor. So on page 26, where we find 51% of Sunnis and 31% of Arabs think life was better under Saddam, those numbers represent a huge majority of Sunni Arabs who say yes offset by equally huge majorities of Sunni Kurds and Arab Shiites, respectively, who say no. To get a rough sense of Sunni Arab opinion, check the region-by-region numbers for Anbar province. In this case, we find 78% preferred Saddam to the new government.

The most shocking result? Fully 26% of the people surveyed said they’d had a family member or relative murdered within the past three years. I wonder what a statistician could do with that number re: estimating a total post-invasion death toll. It sounds like it’d be big — Lancet-level big — but it’s hard to gauge total “families” from total population. Population is roughly 26 million; make it 24 million for easy calculation. Assume an average of 12 relatives per family for a total of 2 million families, and assume further that one person from every fourth family is murdered. Total murdered: 500,000. Increase the average to 24 relatives per family and it drops to 250,000. The problem is, “relative” could mean a lot of things given the tribal organization of Arab culture; the average “family” might be 50 members. Plus, there’s bound to be overlap where survey respondents have family members in common and they’re all identifying the same murder victim. I dunno. Any statisticians out there want to try?

Pages 71-74 break down the sample by ethnicity and sect, so pay closest attention to them. The good news is that 61% of Iraqis still identify as “Muslim.” Not “Sunni Muslim” or “Shia Muslim,” just “Muslim.” More surprisingly, among Arabs, Shiites were much more likely to identify by sect: 29% versus only 7% for Sunnis. Even in a mixed city like Baghdad, where you’d expect sectarian consciousness to be sky high, only 8% of the sample is identifying as “Sunni Muslim.” I’m not sure how many Sunnis are still there these days, but my sense is even the remnants are quite a bit more than 8% of the population. So there’s some reason for optimism: identification by sect is still a fringe thing — among Sunnis. Compare, however, the numbers on page 72 for “Sunni Muslims” in Anbar to “Shiite Muslims” in the south of the country. Both of those regions are homogeneous by sect, but there are huge differences: only 16% call themselves Sunni in the Sunni community but more than 50% in some southern provinces call themselves Shiite. I’d be curious to know if that’s a recent thing that sprouted after they won the elections and took control of the government or if it’s an artifact of Saddam-era solidarity in the face of oppression and disenfranchisement. I figure it’s the latter; the obvious analog is black consciousness in America vis-a-vis “white consciousness,” such as it is. Either way, the Shiite awakening makes things a lot easier for Iran.

Finally, did the poll oversample Sunnis and, in particular, Sunni Arabs? Look at column (h) on page 62. Unless I’m reading it wrong, 60% of Sunnis sampled were Arab and only 36% were Kurds. If so, that’s screwy: there are more Sunni Kurds in Iraq than Sunni Arabs. Now look at the first column on page 74, which breaks down the 61% who identified as “Muslim” (instead of “Sunni Muslim” or “Shiite Muslim”). Results: 46% were Sunni and 42% Shiite. 46% of 61% is 28%; add that to the 14% who identified as “Sunni Muslim” on page 71 and we’ve got 42% of respondents identifying as Sunni in one form or another. According to the CIA Factbook, though, the largest estimate of Sunni Muslims in Iraq is 37% of the population. (The smallest estimate is 32%.) If in fact Sunni Arabs were oversampled, it means the actual popularity of the Maliki government is probably a little higher than the survey would indicate. So maybe it is a clear majority who prefer him to Saddam. It’s just that the majority is overwhelmingly Shiite.

Anyway. What’d you do with your Saturday night?

Tags: Islam religion