FACULTY OF SOCIAL SCIENCES

Transcription

1 THE UNIVERSITY OF THE WEST INDIES ST. AUGUSTINE, TRINIDAD AND TOBAGO, WEST INDIES FACULTY OF SOCIAL SCIENCES Department of Political Science Telephone: (868) , Exts , 83058, Fax: (868) CONSTITUTIONAL AFFAIRS AND PARLIAMENTARY STUDIES UNIT CAPSU-CNMG OPINION POLL REPORT Report on the findings of a National Survey to measure public opinion for the 2015 general election in Trinidad and Tobago conducted over the period 21st 24th August, 2015 in Trinidad by Dr. Hamid Ghany Senior Lecturer in Political Science and Coordinator, Constitutional Affairs and Parliamentary Studies Unit (C.A.P.S.U.) University of the West Indies, St. Augustine Hamid.ghany@sta.uwi.edu At the request of the management of the Caribbean New Media Group (CNMG), the Constitutional Affairs and Parliamentary Studies Unit (CAPSU) at the University of the West Indies, St. Augustine Campus was retained to undertake a third national opinion poll so as to provide their viewers and listeners with some sense of where the general election stood in relation to the preferences of the public. There was a specific request to investigate what would be considered the most marginal constituencies, in the assessment of CAPSU, for the September 7th general election. This third opinion poll was conducted over the period 21st 24th August, 2015 under my direction as the lead analyst and my CAPSU colleague, Dr. Maukesh Basdeo, as the second analyst. The questionnaire that was used was pre- tested between myself and Dr. Basdeo as well as some random individuals unconnected to the polling activity. The questionnaires were distributed on Thursday 20th August, 2015 to the field interviewers after a briefing session in which the new focus was explained to them. These field interviewers had previously been given a written guideline on how to conduct a face- to- face poll using the intercept method. The guideline had previously been discussed with them on Wednesday 8 th July, A total of 2,328 questionnaires were administered in Trinidad. The margin of error in Trinidad was calculated at +/ 3%. The face- to- face intercept method employing random sampling techniques was used. Using, once again, the marginality formula that I invented as part of my methodology for the Commonwealth Parliamentary Association (C.P.A.) research project on public perceptions of Parliament in Trinidad and Tobago that I led, I continued the use, from the first poll, of a virtual zoom lens to focus on those constituencies that fell inside the line of 2,300 votes between first and second candidates based on results from the 2010 general election. I used the information for the Revised List of Electors that was made public on 12th August, 2015 which allowed me to calculate the outer limit of marginality (the value of M) so as to list those constituencies which would either fall inside or outside of the value of M. The outer limit

2 2 of marginality with the 12th August 2015 electoral registration data was re- calculated and revised from 3,372 for both the first and second poll and is now expressed as 3,370. Marginality Formula In conducting this third national poll, it was decided to continue the use of parliamentary electoral constituencies as a natural geographical basis for gathering the data. To this end, the official results of the general elections of 24th May, 2010 provided the basis for the choice of constituencies together with the boundary changes contained in the Report of the Elections and Boundaries Commission dated 29th January, I combined that information with my marginality formula and the Revised List of Electors dated 12th August, The use of marginality as a geographical and political basis for seeking opinions would be best owing to the expectation of the presence of relatively equal levels of competing political opinions in the selected constituencies. My marginality formula is designed to calculate an outer limit for marginality in order to prescribe areas where field interviewers should be sent as opposed to predicting electoral outcomes. The symbols for this formula are M for marginality and D for the mid- point spread of the deposit level for the largest (L) and smallest (S) constituencies if everyone were to vote. This formula applies to electoral systems that use the first past- the- post method where the retention of the deposit of every candidate is calculated at one- eighth of the total votes cast in a constituency. The deposit is a fundamental statistical benchmark and can serve as a guide to determine where a survey should be conducted if marginality is the key element in the search for truly mixed political opinions because in marginal constituencies the first and second candidates will always save their deposits owing to the fact that marginality always tends to move closer to zero as the difference between the first and second candidates. The constituencies with the largest electorate (L) and the smallest electorate (S) must be divided by 8 and then their difference divided by 2 in order to calculate D. Therefore, the formula is : (L/8 S/8) / 2 = D Once D is determined, then the statistical formula for calculating the outer limit of marginality (M) is : {(L/8) D = (S/8) + D} = M. Using my marginality formula with the 2015 Revised List of Electors, the following emerged : L = Toco- Sangre Grande 30,148 / 8 = S = Port- of- Spain South 23,777 / 8 = = / 2 = = 3, = 3, M = 3, Under the circumstances, M was rounded down to the nearest whole number for purposes of this survey, so that M = > 3,370. Marginality When the value of M (> 3,370) was applied to the results of the 2010 general election on the basis of measuring the difference between the first and second candidates in all constituencies,

3 3 the following marginal constituencies in Trinidad emerged with a difference between first and second that was less than the outer limit of marginality (> 3,370): Arima (364) (2010 turnout 62.74%) (leaning COP) Diego Martin North East (463) (2010 turnout 60.80%)(leaning PNM) Point Fortin (655) (2010 turnout 68.63%)(leaning PNM) Toco/Sangre Grande (700) (2010 turnout 66.19%)(leaning UNC) Diego Martin Central (993) (2010 turnout 62.86%)(leaning PNM) Lopinot / Bon Air West (1,003) (2010 turnout 71.82%)(leaning COP) Diego Martin West (1,046) (2010 turnout 60.81%)(leaning PNM) La Horquetta / Talparo (1,079) (2010 turnout 68.80%)(leaning UNC) San Fernando West (1,251) (2010 turnout 72.80%)(leaning COP) D Abadie / O Meara (1,343) (2010 turnout 69.65%)(leaning COP) Tunapuna (2,297) (2010 turnout 74.08%)(leaning COP) La Brea (2,564) (2010 turnout 71.07%)(leaning PNM) Port- of- Spain South (2,770) (2010 turnout 53.56%)(leaning PNM) St. Ann s East (2,776) (2010 turnout 60.38%)(leaning PNM) Moruga / Tableland (2,947) (2010 turnout 77.25%)(leaning UNC) Port- of- Spain North / St. Ann s West (2,968) (2010 turnout 57.63%)(leaning PNM) St. Joseph (3,057) (2010 turnout 72.12%)(now leaning PNM since 2013 by election) When the boundary changes recommended by the Elections and Boundaries Commission in their Report dated 29th January, 2014 and approved by the House of Representatives are factored in, the following revised marginality sequence emerged with adjusted first over second place figures : Arima (364) (2010 turnout 62.74%) (leaning COP) Diego Martin North East (was 463 now 618) (2010 turnout 60.80%)(leaning PNM) Point Fortin (655) (2010 turnout 68.63%)(leaning PNM)

4 4 Toco/Sangre Grande (700) (2010 turnout 66.19%)(leaning UNC) Diego Martin Central (was 993 now 832) (2010 turnout 62.86%)(leaning PNM) Lopinot / Bon Air West (1,003) (2010 turnout 71.82%)(leaning COP) Diego Martin West (was 1,046 now 1,269) (2010 turnout 60.81%)(leaning PNM) La Horquetta / Talparo (1,079) (2010 turnout 68.80%)(leaning UNC) San Fernando West (was 1,251 now 1,084) (2010 turnout 72.80%)(leaning COP) D Abadie / O Meara (1,343) (2010 turnout 69.65%)(leaning COP) Tunapuna (2,297) (2010 turnout 74.08%)(leaning COP) La Brea (2,564) (2010 turnout 71.07%)(leaning PNM) San Fernando East (was 3,627 now 2,610) (boundaries changed for 2015) (leaning PNM) Port- of- Spain South (was 2,770 now 2,813) (2010 turnout 53.56%)(leaning PNM) St. Ann s East (2,776) (2010 turnout 60.38%)(leaning PNM) Moruga / Tableland (2,947) (2010 turnout 77.25%)(leaning UNC) Port- of- Spain North / St. Ann s West (2,968 now 2,751) (2010 turnout 57.63%)(leaning PNM) St. Joseph (was 3,057) (2010 turnout 72.12%)(now 780 leaning PNM from UNC since 2013 bye- election)(bye- election turnout 52.99%) These marginal constituencies emerged on the basis of using the difference between the first and second candidates in the results of the 2010 general election and then re- calculating the differences between first and second candidates on the basis of boundary changes implemented by the EBC for the 2015 general election. On 21st April, 2010, five political organizations (the United National Congress, the Congress of the People, the National Joint Action Committee, the Movement for Social Justice and the Tobago Organization of the People) signed a document establishing the People s Partnership in which they undertook not to contest against each other in the forty- one constituencies. This partnership still exists minus the Movement for Social Justice, therefore, this opinion poll, as with the first two, used the classification PP as the option that included the remaining members of the People s Partnership. As a consequence, the following constituencies in Trinidad can be considered marginal in relation to the 2015 general election based on my marginality formula, when combined with revised list of electors and revised boundary information as well as the results of the bye- election in 2013 in St. Joseph, owing to the fact that their projected marginality based on 2010 (and 2013) electoral data fell inside of 3,370 :

5 5 Arima (364) (2010 turnout 62.74%) (leaning COP) *Diego Martin North East (now 618 was 463) (2010 turnout 60.80%)(leaning PNM) Point Fortin (655) (2010 turnout 68.63%)(leaning PNM) Toco/Sangre Grande (700) (2010 turnout 66.19%)(leaning UNC) +St. Joseph (now 780 was 3,057) (2010 turnout 72.12%)(2013 turnout 52.99%)(leaning PNM since 2013 bye- election after being UNC in 2010 general election) *Diego Martin Central (now 832 was 993) (2010 turnout 62.86%)(leaning PNM) Lopinot / Bon Air West (1,003) (2010 turnout 71.82%)(leaning COP) La Horquetta / Talparo (1,079) (2010 turnout 68.80%)(leaning UNC) *San Fernando West (now 1,084 was 1,251) (2010 turnout 72.80%)(leaning COP) *Diego Martin West (now 1,269 was 1,046) (2010 turnout 60.81%)(leaning PNM) D Abadie / O Meara (1,343) (2010 turnout 69.65%)(leaning COP) Tunapuna (2,297) (2010 turnout 74.08%)(leaning COP) La Brea (2,564) (2010 turnout 71.07%)(leaning PNM) *San Fernando East (now 2,610 was 3,627) (boundaries changed for 2015) (leaning PNM) *Port- of- Spain North / St. Ann s West (now 2,751 was 2,968) (2010 turnout 57.63%)(leaning PNM) *Port- of- Spain South (now 2,813 was 2,770) (2010 turnout 53.56%)(leaning PNM) St. Ann s East (2,776) (2010 turnout 60.38%)(leaning PNM) Moruga / Tableland (2,947) (2010 turnout 77.25%)(leaning UNC) * Denotes boundary changes and recalculated electoral data. + Denotes bye- election with revised electoral data. The application of a Zoom Lens to Reduce the Outer Limit of Marginality In the previous two polls, in order to ensure that there was the ability to focus on those constituencies that were the closest to call in respect of their marginality, a decision was taken to systematically reduce the outer limit of marginality from 3,372 (as it was a that time before

6 the publication of the Revised List of Electors on 12th August, 2015) to 2,300 in order to focus on those constituencies that would fall within this range. Based on the recalculation of the results of the general election held on 24th May, 2010 in relation to the new boundaries contained in the Report of the Elections and Boundaries Commission dated 29th January, 2014 and the bye- election data from 2013, the most marginal constituencies using the value of M = >2,300 were calculated as follows : Arima (364) (2010 turnout 62.74%) (leaning COP) *Diego Martin North East (now 618 was 463) (2010 turnout 60.80%)(leaning PNM) Point Fortin (655) (2010 turnout 68.63%)(leaning PNM) Toco/Sangre Grande (700) (2010 turnout 66.19%)(leaning UNC) +St. Joseph (now 780 was 3,057) (2010 turnout 72.12%)(2013 turnout 52.99%)(leaning PNM since 2013 bye- election after being UNC in 2010 general election) *Diego Martin Central (now 832 was 993) (2010 turnout 62.86%)(leaning PNM) Lopinot / Bon Air West (1,003) (2010 turnout 71.82%)(leaning COP) La Horquetta / Talparo (1,079) (2010 turnout 68.80%)(leaning UNC) *San Fernando West (now 1,084 was 1,251) (2010 turnout 72.80%)(leaning COP) *Diego Martin West (now 1,269 was 1,046) (2010 turnout 60.81%)(leaning PNM) D Abadie / O Meara (1,343) (2010 turnout 69.65%)(leaning COP) Tunapuna (2,297) (2010 turnout 74.08%)(leaning COP) These constituencies were polled on the basis of using, as far as possible, the strongest PNM and PP polling divisions respectively in each case in order to seek balanced views across the board. At the same time, by reducing the number of constituencies on a scientific basis using a zoom lens- value of M (>2,300) which fell inside the actual value of M (>3,372), there was scientific certainty that those constituencies selected were worthy of closer scrutiny given their closer proximity to a value of 0 which is the ultimate marginality figure. In doing this, no judgement calls were made in respect of possible outcomes in the remaining six constituencies that fell between 2,301 and 3,372 which constituted the cluster of constituencies that did not fall within the zoom- lens approach adopted for the first two polls. Further Zoom Lens Adjustment Having completed the first two polls, it became apparent that there were six constituencies that would require closer attention because of the information derived from the field that confirmed the existence of a possible closer outcome. 6

7 The six constituencies that demonstrated those characteristics were (i) La Horquetta/Talparo, (ii) Point Fortin, (iii) San Fernando West, (iv) St. Joseph, (v) Toco/Sangre Grande and (vi) Tunapuna. In eliminating constituencies that were not part of the list of constituencies that did not fall into the zone of marginality from the first two polls, namely Chaguanas West, Tobago East and Tobago West, it became apparent that Chaguanas West was leaning heavily in favour of the People s Partnership and Tobago East and West were leaning heavily in favour of the People s National Movement. Accordingly, those three constituencies were not polled in this third survey as the focus of the survey was on the most likely closest outcomes based on polling data derived from the first two polls. The questionnaires that would have been administered in those constituencies were redeployed into the six constituencies listed above on the basis that each constituency would have a minimum of 384 questionnaires per constituency. The margin of error for each constituency was calculated on the basis of using the figures listed for these constituencies on the web site of the Elections and Boundaries Commission based on the Revised List of Electors released on 12 th August, The margin of error for each constituency is as follows : (i) La Horquetta/Talparo 26,026 (384 questionnaires)(moe +/- 4.96%) (ii) Point Fortin 25,505 (384 questionnaires)(moe +/- 4.96%) (iii) San Fernando West 24,492 (384 questionnaires)(moe +/- 4.96%) (iv) St. Joseph 27,651 (393 questionnaires)(moe +/- 4.91%) (v) Toco/Sangre Grande 30,148 (397 questionnaires)(moe +/- 4.89%) 7 (vi) Tunapuna 26,650 (397 questionnaires)(moe +/- 4.96%) The margin of error for the national survey based on the cluster of these six constituencies together is +/- 3.0%. Constituencies and Polling Divisions for Survey All six of these constituencies fell inside of the revised value of M (<3,370). The polling divisions were chosen on the basis of seeking to balance the two strongest PNM polling divisions against the two strongest PP polling divisions and including two marginal polling divisions based on the 2010 general election results and any bye- election results that were available. Where marginality could not be found, the three strongest PNM polling divisions were polled alongside the three strongest PP polling divisions. The survey was conducted in the following constituencies and polling divisions using the addresses for these polling divisions as contained in the official report on the general elections for 2010 in Trinidad to determine actual geographical locations :

8 8 POLLING DIVISION LOCATIONS FOR SURVEY 1. LA HORQUETTA TALPARO 2055 San Rafael R.C. School, Tumpuna Road, San Rafael 2059 La Horquetta South Govt. Primary School, LP. # 54, Gladiolus Crescent, La Horquetta 2062 La Horquetta Regional Complex, Marjorie Padmore Avenue, La Horquetta 2063 La Horquetta RC Church Parish Hall, Uriah Buzz Butler Avenue, Phase III, La Horquetta 2665 Todd s Road RC School, Fletcher s Road, Todd s Road 2680 Las Lomas # 2 Govt. School, Caroni South Bank Road, Las Lomas # 2 2. POINT FORTIN 4696 Guapo Govt. Primary School, Southern Main Road, Guapo 4760 Egypt Government Primary School, Main Road, Egypt Village, Point Fortin 4790 Cap- de- Ville Community Centre, Erin Road, Cap- de- Ville 4800 Chatham Government School, Chatham Road, Chatham 4810 Granville RC School, Syfoo Trace, Granville 4815 Cedros AC School, Southern Main Road, Bonasse Village 3. SAN FERNANDO WEST 3786 La Romaine RC School, Victoria Street Extension (Opposite Health Centre), La Romain 3879 San Fernando Central Secondary School, Todd Street, San Fernando 3972 Marabella United Friendly Society Hall, 4 Prevatt Street, Marabella 4101 Presentation College, Carib Street, San Fernando 4125 San Fernando Methodist Primary School, Mon Chagrin Street, San Fernando 4151 San Fernando Girls Anglican School, Pouchet Street off Rushworth Street, San Fernando 4. ST. JOSEPH 1041 Mt. Hope Secondary School, Maingot Street, Mt. Hope Mt. Hope Secondary School, Maingot Street, Mt. Hope 1055 Trinidad Harmony Friendly Society Hall, 5 Broome Street North, Petit Bourg, San Juan 1470 Aranguez Hindu School, Chootoo Street, Aranguez 1500 Bamboo Grove Presbyterian School, Bamboo Settlement No. 1, Uriah Butler Highway 1506 Cipriani College of Labour and Co- operative Studies, Churchill Roosevelt Highway, Valsayn 5. TOCO SANGRE GRANDE 2145 Toco Secondary School, Galera Road, Toco 2175 North Oropouche RC School, Toco Road, Vega de Oropouche 2180 North Eastern College, Graham Trace, Ojoe Road, Sangre Grande 2185 Valencia South Government Primary School, Alexander Street, Valencia 2235 Sangre Grande Cultural Centre, Sukram Street, Sangre Grande 2242 Sangre Grande Secondary School, Graham Trace, Sangre Grande 6. TUNAPUNA 1620 YWCA Auditorium, 68 Gordon Street, St. Augustine 1625 St. Benedict s RC School, LP. # 34, St. John s Road, St. Augustine 1690 Tunapuna Secondary School, Taylor Street, Tunapuna 1706 El Dorado West Secondary School, Target Street, El Dorado

9 El Dorado East Secondary School, Karamath Street, El Dorado 1880 Maracas Presbyterian School, 52 Acono Road, Maracas, St. Joseph SURVEY REPORT - TRINIDAD The breakdown of the survey is as follows: 1. Gender Male 38.0% Female 62.0% 2. Ethnicity Afro 39.2% Indo 38.3% Mixed 21.7% Not stated 0.8%

10 10 3. Age Group % % % % % 70 and over 4.9% Not stated 0.5% 4. Employment status Employed 36.6% Unemployed 8.0% Professional 6.0% Housewife 10.1% Retired / pensioner 11.3% Self- employed 14.1% Temporary worker 4.6% Student 8.9% Other 0.1% Not stated 0.2%

11 5. Religion Roman Catholic 24.2% Anglican 9.4% 7 th Day Adventist 6.7% Hindu 20.4% Presbyterian 10.9% Methodist 3.8% Pentecostal 16.5% Muslim 5.4% Baptist 1.2% Other 1.5% 11

12 12 6. In casting your vote, what factor will influence you the most? Political Leader 17.3% Candidate in your area % Party loyalty 18.4% Party Manifesto 18.3% Need to retain the government 10.3% Need for change 16.8% Don t know 1.5% Not sure 2.6% Other 0.5% Not stated 0.1% This question was designed to test for voter motivation. 7. Do you believe that you must be a party member or you can just support a party in order to stand for election as a party s candidate? Party member 48.0% Support the party 38.5% Not sure 10.1% Don t know 3.4%

13 13 This question was designed to test opinions on party membership or party support for candidates. 8. Do you believe that foreign guests of a political party should come here to criticize local public officials on a political platform? Yes 16.0% No 62.6% Not sure 15.3% Don t know 6.0% Not stated 0.1% This question is designed to test opinions on a recent controversy about foreign guests criticising public officials on a party platform. 9. Do you believe that the Government should have built a highway from San Fernando to Point Fortin? Yes 52.6% No 28.3% Not sure 13.6% Don t know 5.5% Not stated 0.1%

14 14 This question was designed to test opinions on a controversial public policy project. 10. Do you believe that the Government should build hospitals in Arima, Couva and Point Fortin? Yes 70.5% No 23.5% Not sure 4.8% Don t know 1.2% This question was designed to test public opinion on a major public policy initiative. 11. Do you believe that the Government s laptop programme for secondary school students should be continued or discontinued? Continued 69.5% Discontinued 14.8% Not sure 13.4% Don t know 4.0% Not stated 0.1%

15 15 This question was designed to test public opinion on a major public policy project. 12. Do you believe that the 2009 property tax should be re- introduced? Yes 16.6% No 51.8% Not sure 27.5% Don t know 4.0% Not stated 0.1% This question was designed to test public opinion on a controversial public policy initiative. 13. Do you support the presence of CARICOM and Commonwealth Observer Missions to monitor the September 7th general election? Yes 68.1% No 13.0% Not sure 15.7% Don t know 3.1%

16 16 This question was designed to test public opinion on a general election initiative. 14. Do you believe that the September 7th general election will be free and fair? Yes 53.7% No 21.8% Not sure 20.4% Don t know 3.9% Not stated 0.1% This question was designed to test public opinion on perceptions of the electoral process. 15. If a general election were to be held tomorrow, who would you like to see as the Prime Minister? Kamla Persad- Bissessar 42.4% Keith Rowley 35.3% Jack Warner 5.5% Not sure 12.6% Don t know 4.0% Other 0.2%

17 17 This question was designed to test public opinion on preferences for Prime Minister. 16. If a general election were to be held tomorrow, which group or party would you vote for? PNM 35.4% PP 42.3% ILP 5.4% Not sure 7.5% Don t know 9.1% Other 0.3% This question was designed to test public opinion on party/group electoral preferences.

18 18 SURVEY REPORT LA HORQUETTA/TALPARO (MARGIN OF ERROR +/- 4.96%) 16. If a general election were to be held tomorrow, which group or party would you vote for? PNM 36.4% PP 43.6% ILP 4.2% Not sure 3.4% Don t know 12.5% SURVEY REPORT POINT FORTIN (MARGIN OF ERROR +/- 4.96%) 16. If a general election were to be held tomorrow, which group or party would you vote for? PNM 35.2% PP 48.2% ILP 4.4% Not sure 4.2% Don t know 7.0% None 1.0%

19 19 SURVEY REPORT SAN FERNANDO WEST (MARGIN OF ERROR +/- 4.96%) 16. If a general election were to be held tomorrow, which group or party would you vote for? PNM 40.9% PP 37.0% ILP 7.3% Not sure 6.5% Don t know 7.6% None 0.8% SURVEY REPORT ST. JOSEPH (MARGIN OF ERROR +/- 4.91%) 16. If a general election were to be held tomorrow, which group or party would you vote for? PNM 31.6% PP 44.3% ILP 7.4% Not sure 6.1% Don t know 10.7%

20 20 SURVEY REPORT TOCO/SANGRE GRANDE (MARGIN OF ERROR +/- 4.89%) 16. If a general election were to be held tomorrow, which group or party would you vote for? PNM 39.5% PP 43.3% ILP 2.0% Not sure 6.3% Don t know 8.8%

21 21 SURVEY REPORT TUNAPUNA (MARGIN OF ERROR +/- 4.96%) 16. If a general election were to be held tomorrow, which group or party would you vote for? PNM 28.8% PP 37.1% ILP 7.0% Not sure 18.4% Don t know 8.3% Not stated 0.3% SUMMARY ANALYSIS The People s Partnership continues to maintain its lead that was previously shown in the two earlier polls that were conducted during the period 10 th 13 th July, 2015 and 31 st July 3 rd August, In this poll, which had a revised list of constituencies, that placed the focus squarely on the six most marginal constituencies, the People s Partnership is leading in five of the six constituencies (La Horquetta/Talparo, Point Fortin, St. Joseph, Toco/Sangre Grande and Tunapuna), while the PNM is leading in one (San Fernando West). Based on the margin of error that has been calculated for each of these six constituencies, the following calls can be made in respect of each one : La Horquetta/Talparo leaning PP Point Fortin leaning PP San Fernando West too close to call (inside the margin of error) St. Joseph leaning PP Toco/Sangre Grande too close to call (inside the margin of error) Tunapuna leaning PP CAPSU UWI 28 th August, 2015