Chatham House rules for election fraud
Someone has put up a $10,000 reward for anyone who can prove that the Iranian election was “stolen”, but I don’t think that Chatham House (CH) is likely to win. A Chatham House report prepared by political scientist Ali Ansari claims to have statistical evidence that the election was rigged. I had not planned to write on non-economic issues here but this one qualifies because it deals with numbers.
I have not had a chance to examine the report closely, but here are a few of its claims that do not inspire confidence. Three key points about the key claims made in the CH report:
- The CH report claims that the fact that the variation in participation across provinces has dropped is evidence of fraud. Anyone familiar with elementary statistics knows that the standard deviation of any variable limited to 100% from above would drop as if its mean increases. (At the limit, when the mean is 100%, the SD would be zero!) So, because the participation rate increased by about 35%, it is hardly surprising that the SD fell by 23%.
- CH authors write: “If Ahmadinejad’s victory was primarily caused by the increase in voter turnout, one would expect the data to show that the provinces where there was the greatest ‘swing’ in support towards Ahmadinejad would also be the provinces with the greatest increase in voter turnout. This is not the case.” This argument falls apart the moment you notice that the “if” part is a suspect assumption. Many observers have noted that liberal voters who did not vote in the last election voted this time. The surge in participation could have favored Mousavi.
- Participation in excess of 100% in two provinces. This is not surprising for Mazandaran, where many Tehranis and other well-to-do middle class voters (who most likely voted for Mousavi) were vacationing. The CH report labels Mazandaran as a conservative province but I had no idea that it was conservative. Yazd’s participation in excess of 100% is more of a problem, though I am again not sure about the label conservative (Moussavi won the city of Yazd, which is Khatami’s territory). Now that data is available by polling station, a more careful detailed is possible.
- The CH report rejects the assumption that the rural vote is conservative. I am not sure who has made such a claim. I happen to think that it is more prone to populist instincts, which is not the same as conservative. It responds to giveaways in the short run.
- Finally, superficial comparisons between the 2005 and 2009 elections ignore the fact that Ahmadinejad of 2005 was an unknown while in 2009 he was portrayed as a “man of the people” who had distributed benefits to the poor and who was brave enough to call influential clergy corrupt on television. The two elections are different because Ahmadinejad’s opponents and supporters are different, as are the candidates, including Ahmadinejad himself.
In conclusion, given the way the results have been released and the events that have transpired since, it is hard to blame anyone for doubting the official results. I am surprised, however, to find many intelligent people absolutely convinced that there was massive fraud. Perhaps they have better evidence than CH has provided. I take it for granted that in any developing country with a brief history of voting, especially one with a self-righteous incumbent, election outcomes should be taken with a grain (sometimes a whole bag) of salt. But while skepticism is healthy, certitude without good evidence is not.
Tyranny of numbers 
[…] https://djavad.wordpress.com/2009/06/28/the-chatham-house-rules-election-fraud/ […]
[…] Djavad, “Tyranny of Numbers – Chatham House Rules For Election Fraud,” at https://djavad.wordpress.com/2009/06/28/the-chatham-house-rules-election-fraud/ (June 28, […]
[…] [18] See “Tyranny of Numbers – Chatham House Rules For Election Fraud,” Professor Djavad Salehi-Isfahani, June 28, 2009, https://djavad.wordpress.com/2009/06/28/the-chatham-house-rules-election-fraud/. […]
I feel there was massive fraud in this election, but I can not prove it. I think that with the protest afterwards, and huge numbers that took to the streets despite great personal danger, a rational person can have no doubts.
Very good of you to admit that you cannot prove massive fraud. M\For most people no proof is necessary, which says something about how Iranians think about politics and also about the level of trust between the government and the people. In any case, with all that has happened since these charges were first aired, the question of fraud is fast becoming moot.
Guardian Council released their election report with responses to the complaints.
Your thoughts and analysis would be appreciated.
http://www.shora-gc.ir/portal/Home/ShowPage.aspx?Object=News&CategoryID=4d425fd9-748e-4826-8378-24118396b087&LayoutID=7952d93c-6e32-4abc-8026-eb54e465f88a&ID=4086d4f9-0a6d-4f02-a597-2bdced99f99e
Thanks for the link to the GC report. I have little to say on this matter but, glancing the report, I would love to see a sober analysis of its conclusions from the offices of the three candidates.
Hello Sir
Dr. Walter Mebane from Michigan university did another analysis on data and I would like to ask do you think that these analysis are not good enough too?
Click to access note14jun2009.pdf
thanks a lot
Hope you reply me.
Thanks for the link. I have seen but not read Mebane’s study. It is definitely technically more sohpisticated than the CH report. I am doubtful that a statistical study can produce the smoking gun that people are seeing everywhere. But this type of research is important to dig up anomalies.
The Mebane study relies on Benford’s Law to conclude there was fraud but according to the Carter Center which has monitored more than 70 elections around the world:
“In short, Benford’s Law does not generally apply to electoral data and even in cases where we suspect that it might apply, we find that it does not. All in all, Benford’s Law seems like a very weak instrument for detecting voting fraud. There are many reasons to believe that it does not apply to electoral data, and empirical tests suggest that deviations from the law are not necessarily indicative of fraud.”
Thank you for the information. I am very dubious of claims about uncovering voting irregularity from voting data. They never make their assumptions very clear. For example, those who compare candidate voted in 2005 and 2009 assume that candidate policies in between do not affect voting behavior, which is a strong assumption.
They did not set to prove anything just raised questions. Obviously a huge voter swing is not impossible just improbable. You would find it very hard to convince Iranians that those who boycotted the election in 2005 and turned out in mass for 2009 all voted for AhmadiNejad and that is part of the vote he would have needed to get his 24 million vote.
Remember that Mousavi’s management of war time economy is exemplary in retrospect and many Iranians refer to those times as the “golden” times and many villages may still remember that it was during his time that they got the basic water, electricity and telephone services not to mention the roads he built. So he has his own “populist” credentials.
I am not sure what has happened can b e called a voter swing. This term is useful when there are political parties with well defined programs. I don’t think there was a voter swing when the so-called liberal vote for Khatami (70% plus) shifted to AN in 2005. As I said in my post, AN of 2005 and 2009 are two different candidates.
About Mousavi’s populist credentials, he said he was different and made no attempt to travel to small towns and rural areas (had no time, either). So, his message, which was not populist at all (which is why I liked it), was for reinstating the legal paths, especially in budget allocations and planning. The populist message of AN was that the system was stacked against the poor and had to be dismantled.
I agree. Moreover, I think most of assumptions in CH’s report can be rejected easily based on relatively reliable opinion polls before election like TFT’s (by an international polling organization, 3 weeks before election) and ISPA’s (close to reformists, one week before election and 2 days before election). But Yazd is the province that one should search for possible fraud since as far as I looked at them there is considerable difference between the polls and the results.
What was shocking for me was the following fact one of my friends found in ISPA’s poll (1 week before election poll), the second choice of those who were going to vote for Mousavi:
18.5% Karroubi, Rezaie 6.8%, 31.4% Ahmadinejad!
(The result of ISPA’s poll if you have not seen it was: A 55.1%, M 30%, K 3.8%, R 2.1%.)
Their poll in Ordibehesht gave Ahmadinejad 65%.