ST. LUCIA ELECTIONS DECEMBER 11; COMPETING

C O N F I D E N T I A L BRIDGETOWN 002027 SIPDIS SIPDIS DEPT FOR WHA/CAR AND INR/IAA SOUTHCOM ALSO FOR POLAD E.O. 12958: DECL: 11/17/2016 TAGS: PGOV PREL PINR ST XL SUBJECT: ST. LUCIA ELECTIONS DECEMBER 11; COMPETING POLLSTERS CONFLICT IN PREDICTING VICTORY REF: BRIDGETOWN 1946 Classified By: DCM Mary Ellen T. Gilroy for reasons 1.4(b) and (d). ¶1. (C) SUMMARY: Prime Minister Anthony announced that St. Lucian elections will be held on December 11. Two key polls have predicted conflicting outcomes for the elections. The track record of the pollster predicting that the ruling party will retain its majority in Parliament seems to be the more reliable. END SUMMARY. ¶2. (U) On November 16, St. Lucia's PM Kenny Anthony announced that parliamentary elections will be held December 11, 2006. If PM Anthony's St. Lucian Labour Party (SLP) wins, it will mean a precedent-making third term for the ruling party. Previous to its current administration (1997-present),the SLP held power from 1979-82. PM Anthony pushed the date for this election to the limit: constitutionally, elections should be held every five years. The last elections took place December 3, 2001. Dueling Pollsters -------------- ¶3. (U) Two polls, financed respectively by the two major political parties, reached contradictory predictions of the outcome of the upcoming elections. The Caribbean Development Research Services, Inc. (CADRES),retained by the ruling St. Lucian Labour Party (SLP),predicted an SLP victory. According to the CADRES poll conducted in October 2006, 34 percent of those surveyed preferred the SLP, while 28 percent supported the opposition United Workers Party (UWP),with a margin of error of 5 percent at the national level. The polling results of an American firm hired by the UWP showed that the challengers were ahead in 11 of 17 districts, with two more leaning toward the UWP. ¶4. (U) In the CADRES poll, 37 percent of those surveyed refused to say which party they support and were labeled "uncertain voters." To bolster its prediction of an SLP victory, CADRES noted that, when comparing this year's data to that of a November 2005 survey, voters' references reflect a 10 percent upswing in favor of the SLP against an 8 percent drop in UWP popularity. Furthermore, 29 percent of the "uncertain voters" placed greater confidence in the SLP to lead St. Lucia, while 18 percent were more confident in the UWP's ability. Finally, those surveyed were asked their preferred choice for Prime Minister, with 47 percent choosing SLP leader and standing Prime Minister Anthony, while only 29 percent supported veteran UWP leader (and former Prime Minister, 1964-79 and 1982-96) Sir John Compton. This outcome is a notable change from the 2005 data in which only 39 percent preferred PM Anthony to 40 percent for Compton. To date, CADRES claims to have achieved 100 percent accuracy in pre-election polling in the Eastern Caribbean. ¶5. (C) The UWP hired an American firm to conduct a poll in late October 2006. Although post has not yet received a copy of the results, a UWP figure reported to PolOff that the methodology of this poll focused more on individual districts than the CADRES poll. According to the UWP source, the results show that the UWP is clearly ahead in 11 districts, the SLP is ahead in four, with two seats too close to call, but likely to go to the UWP. The UWP-funded poll results also indicated a near dead heat in voter preference for prime minister, with Compton ahead of Anthony by just one or two points. Polling Challenges -------------- ¶6. (C) Political consultant Peter Wickham, head of CADRES, explained to PolOff why district polling tends to be unreliable in Eastern Caribbean islands. According to Wickham, it is unwise to place confidence in results based on district seats because of the difficulty in surveying an appropriate sample size. Wickham explained that to have accurate district data, a sample size of roughly 1000 voters would have to be surveyed in each district, which is challenging in the Eastern Caribbean for two reasons. Because the population of each district is small, too large a sample results in oversampling of the necessary clusters. Second, obtaining accurate results from that many people is neither culturally nor logistically feasible (NFI). Therefore, although elections are won by individual district results, survey data attempting to predict outcomes by district are unreliable, according to Wickham. He further stated that, from his extensive experience as a Caribbean pollster, district polling is not necessary because national polling trends commonly reflect district results. Coupling his statistical knowledge with his personal observations of the St. Lucia campaign, Wickham predicts that the UWP will gain a couple seats, but that the SLP will retain its majority--currently 13 of 17 seats. Comment -------------- ¶7. (C) CADRES's statistical methodology uses formulas the firm developed to account for unique characteristics of Eastern Caribbean populations. Wickham credits his "100 percent success rate" at calling elections to this specially tailored methodology. Three weeks away from election day, CADRES's prediction of an SLP victory seems sound and is consistent with post's analysis (reftel). Furthermore, general momentum, as seen in the press, at political rallies, and in the form of banners and posters around the capital Castries, appears to reflect the ruling party's stronger position. OURISMAN