Relative to the expected levels of support in our model, the actual polling averages were remarkably close to what our models predicted. The average discrepancy between polls and reality is about 1.0%, which is well within the standard error range and quite surprisingly accurate, especially considering the inaccuracy and the paucity of polling in Iran.
|Candidate||Expected Popular Support||Actual Vote Percentage||Error|
|Raisi + Ghalibaf||40.61||38.3||2.3|
We assumed that this support would be matched by the undecided voters who would shift towards Raisi as a result of both Raisi and Ghalibaf merging their support. We overestimated the level of support of Raisi – this is probably because we underestimated the total amount of voters who would have stuck with Ghalibaf and refused to shift their vote to Raisi.
Notwithstanding this result, considering how had very limited polls (only one western polling sources and several, often inaccurate Iranian polls), we were overall extremely successful.
Out of all polls on the Iranian election, the model based off upon the iPPo group were by far the most accurate, and carried the most weighting in the final estimated levels of popular support. It was able to capture the shifting levels of support before and after In contrast, polls provided by Iranian sources such as the ISPA were substantially less accurate, showing a Rouhani with less than 45% of the vote, which was substantially less than his actual and our predicted measures of a Rouhani majority.
In the grander scene of Iranian politics as a whole, Rouhani will be re-elected to be the President of Iran, in an refutation of Iranian conservatives and a stunning success for the reformers. This election was and is of great importance to the political shape of the region as a whole.
1. Obtained by the sum of the popular support for Ghalibaf and Raisi. Estimated vote breakdown: 18.5% for Ghalibaf and 21.8% for Raisi.↩