Monday, June 15, 2009

"Terror Free Tomorrow Poll Did not Predict Ahmadinejad Win" (Juan Cole)

As I have indicated, it looks clear to a wide range of analysts and observers--including a non-expert observer like me--that the official results for Friday's presidential election in Iran are almost certainly bogus. In fact, the more one examines them in detail, the further they seem to go way beyond the bounds of credibility, or even plausibility.

It is only fair to add that not everyone agrees. Even some normally sensible people seem to have taken the official results at face value--including Christopher Dickey, to my surprise. But they don't address the many grounds for being skeptical about the officially announced figures.

Others have suggested that, at the very least, Ahmadinejad has enough popular support (and enough built-in institutional advantages, including the disqualification of most of the potential alternative candidates by the Guardian Council and the backing of the Revolutionary Guard apparatus) that he could have won a fair election even without massive vote-rigging. That possibility doesn't seem inherently implausible, but at this point we'll never know what the outcome of a fair election would have been.

A few analysts have even argued that the official election results look plausible. I confess that I haven't seen any such arguments that I found very solid or convincing. (Most of them simply repeated the point that, after all, Ahmadinejad does have genuine support in some sectors of Iranian society--but what sensible person denies this?) The most serious attempt to make such an argument that I've run across was presented in a Washington Post op-ed by Ken Ballen and Patrick Doherty, based on a telephone poll of Iranians they had conducted three weeks before the election for Terror Free Tomorrow.

In a careful, respectful, but quite devastating analysis (see below), Juan Cole explains why the interpretation offered in their op-ed is fallacious, misleading, and less than plausible. Cole's discussion is clear and straightforward, so it requires no summarizing or explication.

For anyone who isn't entirely convinced by Cole's critique, the Ballen & Doherty op-ed has also been convincingly debunked by Nate Silver (at his legendary FiveThirtyEight blog) and Jon Cohen (at the Washington Post), among others. The two critics I've just mentioned, unlike Cole or me, have considerable expertise in the statistics of polling.

=> The bottom line is that while there is obviously no direct proof (so far) that the official election results were falsified, there are many reasons to find them extremely non-credible ... and, as far as I know, no one has yet offered an especially convincing argument for why we should accept those official results as valid.

--Jeff Weintraub

P.S. Flynt Leverett and Hillary Mann, a pair of foreign-policy "realists" who have long urged that the US should be more accommodating to the Iranian regime, immediately declared: "Ahmadinejad won. Get over it". Most of their piece consists of simple assertions, but the main piece of actual evidence they cite is the Terror Free Tomorrow poll. Since that evidence turns out not to be very compelling, their conclusion seems to rest on preconceptions and wishful thinking more than anything else.

UPDATE 6/18/2009: For elaboration, see also Iran Election Fraud: Mansoor Moaddel on Ballen and Doherty.

=========================
Juan Cole (Informed Comment)
Monday, June 15, 2009
Terror Free Tomorrow Poll Did not Predict Ahmadinejad Win

Noting my skepticism about the announced outcome of Friday's presidential elections in Iran, readers have been asking me what I think about this WaPo op-ed by Ken Ballen and Patrick Doherty pointing out that a scientifically weighted Project for a Terror Free Tomorrow poll in mid-May found Ahmadinejad beating Mir-Hosain Mousavi by a 2 to 1 margin.

I have enormous respect for Ballen, PFTFT and Doherty & the New America Foundation.

But as a mere social historian I would say that the poll actually tends to confirm some of my doubts about the announced electoral tallies.

The poll did not find that Ahmadinejad had majority support. It found that the level of support for the incumbent was 34%, with Mousavi at 14%.

27% said that they were undecided. (Some 22% of respondents are not accounted for by any of the 4 candidates or by the undecided category, and I cannot find an explanation for this. Did they plan to write in for other candidates? A little over a quarter of respondents did say they wanted more choice than they were being given. Update: Some of this 22% refused to answer, others said they did not like any of the candidates. Ahmadinejad is unlikely to have picked up the latter, and Mousavi supporters were more likely to refuse to answer.)

Here's the important point: 60% of the 27% who said they were undecided favored political reform. As Ballen wrote at that time:
' A close examination of our survey results reveals that the race may actually be closer than a first look at the numbers would indicate. More than 60 percent of those who state they don’t know who they will vote for in the Presidential elections reflect individuals who favor political reform and change in the current system.'
That is, supporters of the challenger's principles may not quite have committed to him at that point but were likely leaning to him on the basis of his platform. They were 16% of the sample. This finding suggests that in mid-May, Mousavi may have actually had 30% support.

If Ahmadinejad got all of the other 11% among undecideds, the race would have stood at 45% to 30%.

Ballen noted in May,
'The current mood indicates that none of the candidates will likely pass the 50 percent threshold needed to automatically win; meaning that a second round runoff between the two highest finishers, as things stand, Mr. Ahmadinejad and Mr. Moussavi, is likely.'
That is, based on his polling, Ballen did not expect Ahmadinejad to get to 51%.

In fact, the regime has announced that Ahmadinejad received almost 63% of the vote. So while Ballen's polling does suggest that it was plausible that Ahmadinejad could have won a run-off election against Mousavi, it indicated that Ahmadinejad was unlikely to win a first round.

Moreover, given the PFTFT numbers, all of the undecideds would have had to vote for Ahmadinejad in order for him to get over 60% of the total vote. That outcome seems to me so statistically unlikely as to rate as an impossibility.

Note that the regime is not merely claiming that Ahmadinejad barely avoided a run-off by getting 51% of the vote. They are saying he received nearly two-thirds of the vote. No such outcome was predicted by the PFTFT poll-- quite the opposite.

So my commonsense, non-technical, historian's comment is that the poll may well have been sound, and Ballen's original conclusions may also have been. But the tenor of his WaPo article contradicts the poll in seeming to find a 63% margin of victory for Ahmadinejad plausible on the basis of it.

Particularly puzzling is that he seems to have forgotten his own observation that the race in May was closer than it seemed, since 60% of undecideds identified with reform principles.

Finally, 42% of respondents successfully contacted declined to answer the poll. Since it is much more likely that reformists would be afraid of government reprisal and afraid of talking about their politics than that Ahmadinejad supporters would be, the possibility that declines were disproportionately pro-Mousavi voters is strong. Although Ballen says voters were willing to answer controversial questions on press freedom or voting for the supreme leader, in fact these are vague and general issues. Imagine if a woman was pro-Mousavi and the phone rang when her husband, a pro-Ahmadinejad voter, was present. She might well just hang up rather than risk a domestic squabble. The decline rate strikes me as quite large, and of a sort that might well skew the results toward Ahmadinejad supporters.