(C) POLL DATA CONFIRM POOR SHOWING OF IAF, BUT

S E C R E T SECTION 01 OF 03 AMMAN 000140 SIPDIS NOFORN SIPDIS E.O. 12958: DECL: 12/11/2017 TAGS: KDEM PGOV KISL JO SUBJECT: (C) POLL DATA CONFIRM POOR SHOWING OF IAF, BUT POINT TO IRREGULARITIES IN JORDANIAN ELECTION RESULTS FOR TWO DISTRICTS REF: AMMAN 4430 Classified By: Ambassador David Hale for reasons 1.4 (b) and (d). ¶1. (S/NF) Summary: Tracking and exit polls were conducted by the International Republican Institute-supported Jordan Center for Social Research on Jordan's November 2007 parliamentary elections. The data confirm what the Islamic Action Front denies - that the Islamists' overall poor showing in the polls was genuine, not the result of fraud. However, when those polls are placed alongside the announced (but as yet unofficial) results, there are significant differences in two specific races. In Amman's third and Zarqa's first district, consistent IRI poll data predicted that certain candidates would win - one IAF member and one tribal representative, respectively. Yet in the announced results, minor tribal candidates which the IRI polls consistently placed far down the results table won seats in parliament. There is still debate within IRI on whether or not to release the results publicly, and the potential GoJ reaction to IRI's disaggregation of the results by voters' East Banker or Palestinian origin. End Summary. ¶2. (C) Throughout the period leading up to parliamentary elections in Jordan on November 20, 2007, the International Republican Institute (IRI) conducted polls in Jordan through the Jordan Center for Social Research, who in turn employed a local contractor, Middle East Marketing and Research Consultants. Two tracking polls were conducted between October 25-28 (3010 respondents) and November 8-11 (2961 respondents). These polls were followed up by exit polling on election day (3729 interviews). These results are now being compared to the announced results of the parliamentary elections. Though the new parliament has already convened, the Ministry of Interior has yet to release full, definitive election results. The winners were declared on the basis of what IRI is calling "semi-official" results released the day after the election. ¶3. (C) The main findings of the poll for the most part track with conventional wisdom about the election results. The Islamic Action Front - the Jordanian political wing of the Muslim Brotherhood - fared poorly across the board, garnering approximately half the number of votes it received in 2003. �00A;IRI polling indicates that this was mostly due at least in part to a boycott or exodus of Palestinian-origin voters. Other political parties did even worse, failing to elect even one of their number to parliament. As expected, most voters cited tribal loyalty and the ability to deliver government services as the primary reason for supporting their candidate of choice. In many districts, the tribal vote was split as pre-election maneuvering often failed to produce a single candidate from each tribe. The turnout rate for women tracks with their participation in the municipal elections, but youth turnout was lower than expected. Polls vs. Results -------------- ¶4. (S/NF) IRI polling data were collected in twenty-one of forty-two total districts throughout Jordan. These twenty-one districts were selected by IRI due to their representative nature and general interest to outside observers (i.e., districts in which the Islamic Action Front was slated to do well). The districts were a mix of demographic and political patterns. In eighteen of these races, tracking and exit polls successfully predicted the outcome of races with a great deal of precision. ¶5. (S/NF) In Amman's fourth district, IRI's tracking polls differ widely from its exit polling. Those election day polls, however, are consistent with the announced results. IRI attributes this to a last minute tribal alliance, in which a bloc of voters were pledged to a certain candidate. Paul McCarthy, IRI's resident director in Jordan, calls Amman's fourth district a "unique mix of demographics," and sees the data there as volatile in that there were large swings in favor of different candidates. ¶6. (S/NF) This leaves two districts (Zarqa first and Amman third) where consistent tracking and polling data do not match the semi-official results. IRI conducted exit polls throughout the day, and has data on the "horse race" as it progressed up until the close of polls. In the end, results for the winning candidates in the two questionable districts were markedly different from what voters themselves were consistently telling pollsters. In IRI's words, "there were at least two 'surprise winners' who did not do well in either AMMAN 00000140 002 OF 003 the tracking polls or the exit poll but somehow managed to win a parliamentary seat, thereby displacing a candidate who was predicted to be a 'sure winner'." ¶7. (S/NF) In Zarqa's first district, there is a twenty percent difference between consistent tracking and exit poll information and the Interior Ministry's "semi-official" results. Salameh Al-Ghweiri was consistently marked as the top vote-getter in IRI tracking polls. On election day, data gathered from voters through exit polling confirmed the expectation of tracking polls, with 14.7 percent of the vote going to Al-Ghweiri. Yet in the official results, Al-Ghweiri came in third, losing out on a seat by 158 votes. Al-Ghweiri is not a member of the IAF, but a tribal candidate who was running for re-election. Note: IRI polls gave the IAF candidate in Zarqa's first district a "fair chance," but he was defeated. End Note. Al-Ghweiri was the chairman of the parliament's Administrative Committee, which was in charge of putting together the municipal elections law, among other pieces of legislation. ¶8. (S/NF) The differences between exit polling and announced results in Zarqa's first district changed the results in favor of Fawaz Hamdallah, an unknown tribal candidate who had never held public office before. IRI exit polling put Hamdallah in eleventh place in Zarqa's first district, with 4.2 percent. In official results, Hamdallah placed second, winning a seat in parliament. Even if the three percent margin of error is taken into account, the gap between Al-Ghweiri and Hamdallah is still suspect. ¶9. (C) Despite the polling data and conventional wisdom that picked him as the overall winner, Al-Ghweiri told IRI staff that he would be denied a seat in parliament because his brother Muhammad Musa Al-Ghweiri is the mayor of Zarqa. This is merely a theory, but one that tracks with other theories from the campaign trail, which state that tribes are "allowed" one elected position per cycle (Ref A). ¶10. (S/NF) In Amman's competitive third district, pre-election IRI tracking polls indicated that an IAF candidate, Irhayl Gharaibeh, was "a sure winner." Exit polls put him in third place with 10.6 percent of the vote - enough to win a seat in parliament. Yet according to official results, Gharaibeh came in fourth place. He lost out to Ismail Al-Bustanji, who IRI exit polls put in seventh place with a meager 3.9 percent of the vote. Again, even considering the poll's margin of error, there is a noticeable difference between the reported preference of voters and the official results. ...But Were the Polls Accurate? -------------- ¶11. (C) It is important to note that while results in these two districts did not align with IRI expectations, the majority of the districts in which they conducted their polls turned out as expected. In particular, the generally poor performance of the IAF was evident in IRI polls long before the elections took place. Tracking polls put IAF support at 7.6 percent nationally, and the numbers released so far by the Ministry of Interior show the actual result to be 7.0 percent. ¶12. (C) It is difficult to draw a clear line between results in the two questionable districts and government interference. Any number of factors could have contributed to the gap in results, including sampling errors, untruthfulness on the part of voters, or other statistical problems. Yet the implication of deliberate manipulation in the tabulation phase is clear in these two instances. Release of the Data -------------- ¶13. (S/NF) IRI's Jordan Office Director approached poloffs and later the Ambassador to review IRI's evolving strategy on releasing the data (which have so far not been shared widely). IRI's view is that the overall report confirms many points made by the government about the elections, including the IAF's weakness. They do not believe the release of the data on the two-district discrepancy between exit polling results and the final tally will cause significant heartburn, when presented in the overall context of the report. They are worried, however, about the release of data tracking the choices of Palestinian-origin voters compared to that of East Bankers. They fear the trouble could come not from the results themselves (which, contrary to mainstream views, showed that this year, the IAF had much stronger support AMMAN 00000140 003 OF 003 among East Bankers than among Palestinian-Jordanians),but because the IRI posed this delicate question at all. So that particular finding may not see the light of day. Hale