The Caribbean Development Research Services (CADRES) has challenged the ruling St Lucia Labour Party to publish for public scrutiny, the results of two polls it says it had conducted and to identify the pollsters.
The challenge was issued after the SLP questioned the reliability of a CADRES poll announced earlier in the week which gave insights into the state of the contest between the two major political parties ahead of Monday’s general election.
CADRES said publication of the SLP’s poll results and the names of the pollsters would “allow the public to independently assess the integrity of the respective pollsters”. Failing to do so CADRES said would leave the St Lucian public with no alternative than to conclude that the SLP’s pollsters either do not exist or that their data is a fiction.
The following is the full text of the CADRES statement:
“CADRES notes the SLP’s response to its most recent poll of political opinion in St Lucia and feels compelled to correct several inaccuracies conveyed in this statement that appears more propagandistic than scientific.
“Ironically the SLP statement begins by indicating that it takes election polls seriously and thereafter proceeds to contradict such seriousness by suggesting that the poll confirms “the lead of the Saint Lucia Labour Party over the United Workers Party” which was NOT suggested in the poll at any point.
“The CADRES poll stated clearly that as of last weekend there was a statistical dead-heat between the two parties. This means that the SLP interpretation is flawed. We would therefore also disagree with the assessment of their anonymous pollsters who are reported to have said “the SLP is firmly in sight of an election victory”.
“Instead, we are persuaded that based on a comparison of this May 2016 poll and the one we conducted in October of 2015, the UWP currently has momentum, while the SLP’s support level has remained static since last year.
“The convoluted question pressed in the SLP response confuses voter turnout with promised support and our associated prediction. The SLP attempts to imply a methodological flaw on the part of CADRES. CADRES has NEVER represented that it was the intention of this poll to predict voter turnout. This would require an entirely different methodology.
“We did, however, intend to predict the likely electoral swing using a method which is consistent with that which correctly predicted a number of elections in the Commonwealth Caribbean; namely 2.6% in favour of the St. Lucia Labour Party in 2011; 1% in favour of the Unity Labour Party in 2015; 12% in favour of the People’s National Movement in 2015 and 9% in favour of the Antigua Labour Party in 2014.
“We are therefore equally confident that our prediction of 1.7% in favour of the UWP will hold. This means that the UWP has sufficient momentum to achieve the minimum 2% swing necessary to win the June 6th election.
“The disaggregation of what CADRES refers to as the “Uncertain Voter” (and is more commonly known as the uncommitted voter) informed the assessment of the likely outcome of the four most recent elections in the Caribbean where the CADRES prediction was vindicated.
“This adequately discharges any SLP suggestion that the CADRES analysis is flawed. Regarding the identification of preferred leaders, the SLP here also makes obvious and deliberate errors and omissions, which a cursory reading of their press statement would reveal.
“CADRES contended that the two leaders are at present equally preferred. This is obvious in the context of a poll with a +/- 5% margin of error, where both Dr. Anthony and Mr. Chastenet are separated by 3% points.
“The SLP is, either, not familiar with the concept of the “margin of error” or, perhaps, believes that the St Lucian public is not sophisticated enough to appreciate this concept.
“There is, however, no such technical appreciation necessary to understand our assertion that Mr. Chastanet has doubled his support among “Uncertain Voters” since he held 7% among this group in 2015 and now holds 13%, while Dr. Anthony has slipped from 18% to 7%.
“The final and perhaps most unfortunate aspect of the SLP’s release was the imputation that the track record of CADRES as an organisation and Peter Wickham as its lead pollster’s track was anything other than successful.
“We believe that the St Lucian public is very familiar with our record and would immediately appreciate the extent to which the SLP is not accurately reflecting the historical record of both CADRES and Mr. Wickham.
“CADRES challenges the SLP to deny that it correctly predicted the correct outcome of the elections held over the last three years in St Vincent and the Grenadines, Trinidad and Tobago, Grenada and Antigua.
“Notwithstanding all of the above, CADRES challenges the SLP to identify the Caribbean polling organisation with a track-record that is superior to ours since 1990 and moreover we ask it to state why it has never challenged a CADRES prediction or methodology in St Lucia previously, although we have been involved in the 2001, 2006 and 2011 elections here.
“Finally, CADRES notes the SLP’s claim that they “recently conducted two of its own polls with two different reputable pollsters”. CADRES challenges the SLP to identify these persons and to make their findings publicly available for scrutiny in the same way that CADRES has done.
“Once the SLP’s pollsters are willing to release their findings, CADRES stands ready to engage these pollsters publicly and compare our respective findings and methodologies in a way that will help to enlighten the St Lucian public on the likely outcome of the election.
“This will allow the public to independently assess the integrity of the respective pollsters, namely the CADRES and the SLP’s anonymous pollsters.
“Should the SLP be unable or unwilling to identify its “reputable” pollsters, then the St Lucian public would be left with no other alternative than to conclude that these persons either do not exist or that their data is a fiction.
“This will create a credibility problem for the SLP with the public at a time when credibility is a serious electoral consideration for voters.”