Friday, September 12, 2003

A reply to the WSJ article on the Zogby poll of Iraqi public opinion

debunking > propaganda
[Warning: huge post... I have included updates in brackets: I discovered the whole poll results that clarify some points raised]
The WSJ article, is based on a Zogby - AEM poll (summary here). The article is a prime example of propaganda, and it illustrates the methods used by the establishment press to turn a perfectly good (in the sense of "interesting" not in the sense of good practice) poll and spin it so as to present it as some sort of legitimization of the US occupation through "Iraqi public opinion".

The Zogby Poll's shortcomings
Before I go in detail about the article, I'd like to point out three serious shortcomings of the poll itself:
  • 1. The poll was done in cities. The rural population of Iraq (a good 33% of the population) was apparently not sampled.

  • 2. For some weird reason Baghdad is not included in the cities sampled. This is bizarre and (in the Opinion Journal piece and the Summary at least) unexplained.

  • 3. Despite consulting "Eastern European pollsters about the best way to elicit honest answers from those conditioned to repress their true sentiments" (hah!), the fact remains that this was an American company asking questions in a country occupied by the US army. This is rather certain to influence the answers of those that were asked, in much the same way that, say, an Indonesian polling company running a similar poll in East Timor, during that country's Indonesian occupation would hardly come up with valid poll results (in fact even if the pollsters were 100% professional and not influenced by the Indonesian government it is rather likely that no one would take such a poll seriously... But that's a different issue).

  • Two more points about the poll as it is presented:
  • 1. The percentage of those refusing to answer is not mentioned.

  • 2. There are two cities from the Kurdish dominated North included in the list of just four cities polled (Kirkuk and Mosul). I do not know how the sample was weighted. However it is quite likely that in the North opinions about the occupation (which in the case of the Kurds really is viewed as a possible liberation) are more positive than in the rest of the country, as both of the two main Kurdish parties are supportive of the US presence in Kurdistan. Two out of four is kinda heavy representation for a region that has less than 20% of the total population of Iraq

  • The premise
    After all of this let's see what the Opinion Journal aims to prove:

    A few sentences into the article the author (Karl Zinsmeister) states the following: "some of us who have spent time recently in Iraq--I was an embedded reporter during the war--have been puzzled by the postwar news and media imagery, which is much more negative than what many individuals involved in reconstructing Iraq have been telling us."
    Well if he wants to be really puzzled, he should check out Arab and European media imagery. He should also check out Salam Pax's and Riverbend's blogs for some first person accounts, as well as non-embedded reporters' accounts...
    Early in the article it is already obvious that the author really believes things are good and getting better in Iraq. The rest of the article shows him struggling to validate his preconceptions through the poll.

    The analysis
    1. Iraqis are optimistic. Seven out of 10 say they expect their country and their personal lives will be better five years from now. On both fronts, 32% say things will become much better.
    What a surprise: this is a country that has just come out of a terrible war and 13 years of the most debilitating sanctions ever to have been imposed on a country. People were dying in large numbers of preventable and/or treatable diseases. There must be a feeling that they've reached rock bottom and the only way is up. A population whose majority expected things to be worse than they currently are would be a population of clinically depressed citizens in need of serious anti-depressant medication.
    2. They are nervous about democracy. Asked which is closer to their own view--"Democracy can work well in Iraq," or "Democracy is a Western way of doing things"--five out of 10 said democracy is Western and won't work in Iraq. One in 10 wasn't sure. And four out of 10 said democracy can work in Iraq. This was posed as an either / or question. Note that the possible replies are not logically antithetical: It's not, "Democracy can work well in Iraq" vs. "Democracy won't work well in Iraq". To agree with the second option one must both see democracy as impossible in Iraq AND consider it a western way of doing things. Note also that this "democracy" isn't defined: what exactly do the pollsters mean by democracy? Yet despite the problematic nature of the question itself, this is in general a negative answer as far as the author's agenda is concerned. So he must try to find a silver lining. He does: Sunnis were negative on democracy by more than 2 to 1; but, critically, the majority Shiites were as likely to say democracy would work for Iraqis as not. People age 18-29 are much more rosy about democracy than other Iraqis, and women are significantly more positive than men. Apart from the fact that these are hardly unexpected results (especially about the younger age group), its worth noting that the details about the demographic breakdown of the numbers appear only when they are supportive of the authors opinion throughout the article.
    3. Asked to name one country they would most like Iraq to model its new government on from five possibilities--neighboring, Baathist Syria; neighbor and Islamic monarchy Saudi Arabia; neighbor and Islamist republic Iran; Arab lodestar Egypt; or the U.S.--the most popular model by far was the U.S.
    You must be kidding me... The options here were four dictatorships and a single representative democracy and we're supposed to be surprised that a (relative) majority chose the democracy as a preferred government? The question that would be really interesting would be to include another democratic country in the question (say Germany) and see what the results would be. Still as Zinsmeister mentions it is just a relative majority: The U.S. was preferred as a model by 37% of Iraqis selecting from those five [23.3% of the total mihalis]-- more than Syria, Iran and Egypt put together. Saudi Arabia was in second place at 28% [17.4% of the total]. This is bad news folks. There are only 23.3% (+ a possible 15.4% "other"), a total minority of the sample) that prefer a representative democracy over four different versions of authoritarianism. 17.5% of the sample look up to Saudi Arabia (a despotic monarchy, the most fundamentalist country after Afghanistan) as a role model. This is supposed to be good news...? Yet, again, when some sort of negative result emerges the author digs into the demographic breakdown to salvage something spinable as positive: Younger adults are especially favorable toward the U.S., and Shiites are more admiring than Sunnis. Interestingly, Iraqi Shiites, coreligionists with Iranians, do not admire Iran's Islamist government; the U.S. is six times as popular with them as a model for governance. Again, the fact that younger people have a more positive view of representative democracy is hardly a surprise. And to assume that "correligionist" Arabs would be admiring of Iran (having lost many of their children to the war against the Iranians) is ... uninformed. Oh and the real fundamentalist regime is not Iran (comparatively speaking) but Saudi Arabia, the approval of which among the Shiia remains unstated [update:its 18%!].
    [Update: the US is actually selected by 23.3%, "Other" gets an unmentioned 15.4% and "Not Sure/None" gets a whopping 21.9%]
    4. Our interviewers inquired whether Iraq should have an Islamic government, or instead let all people practice their own religion. Only 33% want an Islamic government; a solid 60% say no. Again the surprise is in that such a large percentage of Iraqis seem to be backing an Islamic government... This was a thoroughly (and violently) secularized society, in which women could walk around the cities without scarves on their heads. This is a surprise only among those for whom "Arab" is synonymous with "islamic fundamentalist". This statistic would be interesting if we had similar numbers for 3 months ago... A vital detail: Shiites (whom Western reporters frequently portray as self-flagellating maniacs) are least receptive to the idea of an Islamic government, saying no by 66% to 27%. It is only among the minority Sunnis that there is interest in a religious state, and they are split evenly on the question. True some western (he means American corporate media) reporters portray the Shiia as self-flagellating maniacs. However other, better informed, western reporters don't .
    5. 57% of Iraqis with an opinion have an unfavorable view of Osama bin Laden, with 41% of those saying it is a very unfavorable view. (Women are especially down on him.) Except in the Sunni triangle (where the limited support that exists for bin Laden is heavily concentrated), negative views of the al Qaeda supremo are actually quite lopsided in all parts of the country.
    Where to begin: Notice the evasion: among "Iraqis with an opinion". This is the first (but not the last) time in the article that this distinction is made. Obviously, without this distinction the percentage of people with an unfavorable view of ObL sinks below 50%, which, as far as spin is concerned, is bad [update: its 47% against, 36% for...]. The author claims that ObL is quite unpopular except in the "Sunni triangle" (I assume he means Ramadi because there are no other polling spots from the Sunni triangle in this report), which is the only part of Iraq he could be very popular, noting his distaste for Shiites in general [update: in Al Ramadi he has a positive 95% approval rating and no one polled disapproved of him]. And the meagerness of the support is never translated into a concrete number in the article... [update 22% favorable and 23% NF&NS]. Again though, if the numbers are truly against him - no surprise: a majority of secular and non-secular Arabs have little in common with the crazy bastard anyway...
    6. The author also manages to act surprised that Iraqis hated Saddam. Stop the Press. Big News.
    7. With this meager evidence (note that nowhere does he mention numbers directly related to the war or the occupation) the author makes a record-breaking leap of logic to claim that: This new evidence on Iraqi opinion suggests the country is manageable. If the small number of militants conducting sabotage and murder inside the country can gradually be eliminated by American troops (this is already happening), then the mass of citizens living along the Tigris-Euphrates Valley are likely to make reasonably sensible use of their new freedom.
    It suggests no such thing of course. There is no question related to the degree of support these militants have among Iraqis, and the questions about the occupation forces that have been asked in the poll, Zinsmeister leaves to discuss after his conclusion. I cringe to ponder what possible meaning the word "responsible" might have in this context...
    8. The bad news: Inchoate anxiety toward the U.S. showed up when we asked Iraqis if they thought the U.S. would help or hurt Iraq over a five-year period. By 50% to 36% they chose hurt over help. Let me also suggest that a heavy Kurdish bias [update: true, Kirkuk is very pro-american and Al Ramadi very anti-american] makes the answers to this question even more damning for the occupation forces. Note that the anxiety is "inchoate" (though none of the other opinions are affected by this). Well, do we follow the numbers to their logical conclusion? For true believers, true ideolepsy demands that we don't, so we write pitiable and unsupported excuses mentioning feminine aversion to war as a mitigating factor (how this has any bearing on the argument is beyond my humble reasoning skills). The ideoleptic seizure comes to a climax with what is by far the most ridiculous of statements in a thoroughly ridiculous piece of opinion writing: Evidence of the comparative gentleness of this war can be seen in our poll. Less than 30% of our sample of Iraqis knew or heard of anyone killed in the spring fighting... First of all the question didn't say "anyone". The question from the Zogby site is, verbatim, the following: "Were any of your own family members, neighbors, or friends killed in this spring's war?" Thus the question concerns being directly affected by someone's death, not witnessing or hearing of a stranger / acquaintance dying...
    So. Let's get this straight: this is a war which lasted for a few weeks, a few weeks, yet between one in four and one in three of a population of 24 million (or at least those sampled), lost someone really close to them. And Baghdad wasn't even among the cities polled! This is evidence for the comparative gentleness of the war? What's the standard? Hiroshima? Dresden? The Huns? This statement transcends partisanship and moves into the realm of the complete pathological failure of rational faculties.
    But wait! there's more! He then compares this number with the results of the following (as stated in the Zogby site) question: Were any of your own family members, neighbors, or friends killed in war or by security forces during the years that Saddam Hussein held power? [emphasis mine]. Which is carried over to the Opinion Journal piece as "...Meanwhile, fully half knew some family member, neighbor or friend who had been killed by Iraqi security forces during the years Saddam held power..." Note that the word in war is omitted (for a reason) from the Journal's article. So, one in two Iraqis had a person close to them die as a result of the Iran-Iraq war (~10 years, ~ 500,000 casualties), the invasion of Kuwait (Unknown number of casualties, quite probably near the 100,000 mark) 12 years of continuous bombarding by the US & UK air-forces, and 25 years of a cruel and murderous regime that killed rather than jailed its political opponents and this compares favorably (for the US) with three weeks of war resulting in one in four Iraqis losing a loved one!?!??? I'm speechless.
    9. The article then concludes: Perhaps the ultimate indication of how comfortable Iraqis are with America's aims in their region came when we asked how long they would like to see American and British forces remain in their country: Six months? One year? Two years or more? Two thirds of those with an opinion urged that the coalition troops should stick around for at least another year. Ah! Those "with an opinion" again [10% don't have an opinion and the Kurdish unease is a big factor...]. Yet the fact that most of the responding Iraqis see the occupation forces as some sort of police force makes sense. This answer makes sense since there was no mention of possible alternatives in the survey question. Not "should the US forces leave and be replaced by a UN / Arab League force", nor "should leave and be replaced with the Iraqi army". Just a choice between lawlessness and occupation. Some choice!

    The questions we don't hear about

    Listed in the Zogby site are these questions, not mentioned in the Opinion piece:
    - Should America and Britain help make sure a fair government is set up in Iraq, or should they just let Iraqis work this out themselves? [The answer is 32% "US and UK should help", 60% "Iraqis alone", the rest "Not Sure"... Yeah, I wonder why no mention was made of this result...]
    and these
    - Over next five years will Iran help, hurt or have no influence on Iraq- [21% "help", 53% "hurt", 25% "No Influence & NS"]
    - Over next five years will Saudi Arabia help, hurt or have no influence on Iraq [61% "help"(!!!), 7.5% "hurt", 32% "NI&NS"]
    - Over next five years will The United Nations help, hurt or have no influence on Iraq [ 50% "help", 18% "hurt" and 25% "no influence"]

    The questions never asked

    A poll can never be exhaustive. However there is a conspicuous absence of questions related to a possible UN presence in Iraq, to Arab League involvement, to approval of the occupation forces' conduct, to an assessment of the occupation leadership, to an assessment of people like Chalabi who are being groomed for leadership, to attitudes towards Israel, to attitudes towards the EU and many, many, more. It's like someone took the trouble of excluding the questions with the potential for more damage to the tarnished image of american policy in the middle east (after all the American Enterprise Magazine is hardly non-partisan).

    All in all: A less than perfect example of the art of spin...

    No comments: