Guardian Me­dia Ltd did an opin­ion poll on Oc­to­ber 12 and 13 in the con­stituen­cy of St Joseph us­ing the in-per­son ran­dom sam­ple in­ter­cept method.A to­tal of 602 ques­tion­naires were ad­min­is­tered and the mar­gin of er­ror was cal­cu­lat­ed at + or – four per cent. The sur­vey was con­duct­ed in 12 polling di­vi­sions:

�2 1,030–Mt D'Or Gov­ern­ment School, LP 22, Mt D'Or Road, Champs Fleurs

�2 1,041–Mt Hope Sec­ondary School, Main­got Street, Mt Hope

�2 1,075–San Juan Boys' RC School, Ceme­tery Street, San Juan

�2 1,470–Aranguez Hin­du School, Chootoo Street, Aranguez

�2 1,485–Mt Lam­bert RC School, 8th Street, Mt Lam­bert

�2 1,490–Mt Lam­bert RC School, 8th Street, Mt Lam­bert

�2 1,496–Cipri­ani Labour Col­lege, Churchill-Roo­sevelt High­way, Val­sayn

�2 1,500–Bam­boo Grove Pres­by­ter­ian School, Bam­boo Set­tle­ment No1, Uri­ah But­ler High­way

�2 1,507–Curepe Com­mu­ni­ty Cen­tre, Salde­nah Terr, off South­ern Main Rd, Curepe

�2 1,510–St Joseph Govt Pri­ma­ry School, Cor Aber­crom­by & Mar­ket St, St Joseph

�2 1,515–St Joseph's Col­lege, 1 Rich­mond Street, St Joseph

�2 1,525–St Joseph Girls' RC Pri­ma­ry School, Rich­mond Street, St Joseph

These polling di­vi­sions were cho­sen on the ba­sis of seek­ing clus­ters of the most mar­gin­al out­comes from the 2010 gen­er­al elec­tion re­sults to­geth­er with the off­set of strong polling di­vi­sions for the PNM and the UNC us­ing the same 2010 re­sults.This method­ol­o­gy pro­duced a ge­o­graph­i­cal spread that was rep­re­sen­ta­tive of the lay­out of the con­stituen­cy and ap­prox­i­mate­ly 50 ques­tion­naires were ad­min­is­tered in each polling di­vi­sion.

Vari­ables

It must be not­ed that there were vari­ables to be con­sid­ered in re­spect of the col­lec­tion of the da­ta over the pe­ri­od Oc­to­ber 12-Oc­to­ber 13.The vari­ables were:

�2 There was a na­tion­al de­bate among po­lit­i­cal par­ties on tele­vi­sion on the evening of Oc­to­ber 10 in which rep­re­sen­ta­tives of four par­ties par­tic­i­pat­ed. The per­for­mances of the var­i­ous rep­re­sen­ta­tives may have in­flu­enced the re­spons­es.

�2 On Oc­to­ber 12, the Prime Min­is­ter of­fi­cial­ly opened the Grand Bazaar in­ter­change at the in­ter­sec­tion of the Churchill-Roo­sevelt High­way and the Uri­ah But­ler High­way. This in­ter­change lies at the south-west­ern end of the St Joseph con­stituen­cy and may have in­flu­enced some of the re­spons­es.

�2 There was on­ly one known can­di­date up to the pe­ri­od of da­ta col­lec­tion and that was the PNM can­di­date.

�2 Trinidad is in the midst of a high lev­el of cam­paign­ing for the lo­cal gov­ern­ment elec­tions that are to be held to­mor­row.

The break­down of the polling da­ta gath­ered is as fol­lows:

1. Gen­der:

Male-57 per cent

Fe­male-43 per cent

2. Eth­nic­i­ty:

Afro–32.2 per cent

In­do-34.1 per cent

Mixed–32.2 per cent

Chi­nese–0.2 per cent

Syr­i­an/Lebanese–0.8 per cent

Not stat­ed–0.5 per cent

3. Age Group:

18-29–12.8 per cent

30-39–37.7 per cent

40-49–10.1 per cent

50-59–14.6 per cent

60-69–20.4 per cent

Over 70–4.3 per cent

4. Em­ploy­ment sta­tus:

Em­ployed–27.9 per cent

Re­tired/pen­sion­er–27.9 per cent

Un­em­ployed–15.4 per cent

Self-em­ployed–9.3 per cent

Pro­fes­sion­al–2.8 per cent

House­wife–10.1 per cent

Stu­dent–6.5 per cent

5. How would you rate the per­for­mance of Her­bert Vol­ney when he served as your MP?

Very good–0.3 per cent

Good–1.5 per cent

Fair–23.8 per cent

Poor–48.2 per cent

Very poor–24.1 per cent

Don't know–1.3 per cent

Not sure–0.8 per cent

The per­for­mance of the for­mer MP for St Joseph, Her­bert Vol­ney, was mea­sured as a start­ing point.There was a very high lev­el of dis­sat­is­fac­tion with his per­for­mance when the re­spons­es of "Poor" and "Very Poor" are com­bined (72.3 per cent).The on­ly oth­er sta­tis­ti­cal­ly sig­nif­i­cant re­sponse to this ques­tion was a 23.8 per cent "Fair" rat­ing.As a con­se­quence, one can as­sume that the is­sue of rep­re­sen­ta­tion will be a fac­tor in the by-elec­tion.

6. Do you be­lieve that he should have re­mained as your MP af­ter he changed his po­lit­i­cal loy­al­ty from UNC to ILP?

Yes–3.3 per cent

No–80.7 per cent

Don't know–12.3 per cent

Not sure–3.7 per cent

There was a very high re­sponse lev­el against the idea of Her­bert Vol­ney seek­ing to re­main as the MP for St Joseph af­ter re­sign­ing from the UNC with an 80.7 per cent rat­ing.The "Don't know/Not sure" cat­e­go­ry ac­count­ed for 16.0 per cent.In large mea­sure, the de­ci­sion to have a by-elec­tion in St Joseph would have been en­dorsed by this re­sponse fac­tor.

7. Are you in favour of a right of re­call for MPs who change their po­lit­i­cal loy­al­ties af­ter they are elect­ed?

Yes–65.4 per cent

No–22.3 per cent

Don't know–8.0 per cent

Not sure–4.3 per cent

Of those sur­veyed, 65.4 per cent favoured the right to re­call MPs for chang­ing their po­lit­i­cal loy­al­ties af­ter they had been elect­ed, while 22.3 per cent were op­posedThe fact that al­most two-thirds of the sur­vey en­dorsed this phi­los­o­phy would sug­gest that the link be­tween par­ty al­le­giance and elec­tion on a par­ty tick­et has a high cor­re­la­tion.

8. Would you sup­port ex­pand­ing the ex­ist­ing right of re­call to in­clude the mea­sure­ment of per­for­mance for MPs?

Yes–99.3 per cent

No–0.5 per cent

Don't know–0.2 per cent

This ques­tion sought to ex­pand the right of re­call to in­clude mea­sure­ment of per­for­mance of MPs be­yond their par­ty loy­al­ties.

Al­most all those polled (99.3 per cent) were in favour of such a re­form be­ing in­tro­duced.

9. How would you rate the per­for­mance of the Peo­ple's Part­ner­ship Gov­ern­ment to date?

Very good–3.0 per cent

Good–23.3 per cent

Fair–27.2 per cent

Poor–36.9 per cent

Very Poor–9.1 per cent

Not sure–0.5 per cent

This ques­tion was de­signed to start a process of mea­sure­ment of per­for­mance of the PP Gov­ern­ment to date. It was ap­par­ent that more peo­ple felt its per­for­mance was "Poor/Very poor" (46.0 per cent) as op­posed to those who felt that it was "Good/Very good" (26.3 per cent). The crit­i­cal sta­tis­tic, there­fore, was the 27.2 per cent "Fair" rat­ing that was earned which would rep­re­sent a body of vot­ers who could be swayed one way or an­oth­er.

10. How would you rate the per­for­mance of the PNM as an op­po­si­tion to date?

Very Good–5.6 per cent

Good–19.3 per cent

Fair–17.3 per cent

Poor–34.2 per cent

Very poor–23.4 per cent

Not sure–0.2 per cent

In the same way that the per­for­mance of the Gov­ern­ment was mea­sured, it was al­so nec­es­sary to mea­sure the per­for­mance of the Op­po­si­tion.It was ap­par­ent that more peo­ple felt its per­for­mance was "Poor/Very Poor" (57.6 per cent) than those who felt that it was "Good/Very good" (24.9 per cent).The crit­i­cal sta­tis­tic was the 17.3 per cent "Fair" re­sponse rate it earned.

Com­par­a­tive­ly speak­ing, the PNM's per­for­mance as an op­po­si­tion had a high­er neg­a­tive rat­ing than the per­for­mance of the Gov­ern­ment (46.0 per cent com­pared to 57.6 per cent).This could make a case for a third par­ty to emerge.

11. Do you be­lieve the ILP can be­come a ma­jor force in the pol­i­tics of T&T?

Yes–34.6 per cent

No–27.6 per cent

Don't know–21.3 per cent

Not sure–16.6 per cent

This ques­tion about the ILP and the per­cep­tion of re­spon­dents in St Joseph was based on the fact that the par­ty had won the Ch­agua­nas West by-elec­tion in Ju­ly.It is ap­par­ent that there is a sub­stan­tial body of opin­ion (34.6 per cent) that thinks that the ILP can be­come a ma­jor force in pol­i­tics, while there is enough doubt in the minds of re­spon­dents (Don't know/Not sure-37.9 per cent) to sug­gest that a con­clud­ed view has not been formed just yet.

That un­sure el­e­ment ap­pears to be de­cid­ing whether there is a fu­ture for the ILP in na­tion­al pol­i­tics, while a clear 27.6 per cent is firm in the view that the ILP will not be a ma­jor force in the pol­i­tics of the coun­try.One does not know the ex­tent to which the per­for­mance of the ILP rep­re­sen­ta­tive in the na­tion­al­ly-tele­vised de­bate may have af­fect­ed these re­spons­es.

12. Does it mat­ter to you that Jack Warn­er be­longed to the Peo­ple's Part­ner­ship Gov­ern­ment for three years be­fore he formed the ILP?

Yes–86.2 per cent

No–6.3 per cent

Don't know–4.8 per cent

Not sure–2.7 per cent

The cen­tre­piece of the ILP cam­paign has been the per­son­al­i­ty of Jack Warn­er. It is a mat­ter of pub­lic record that he was a mem­ber of the Peo­ple's Part­ner­ship Gov­ern­ment be­fore he part­ed ways with it of­fi­cial­ly on Ju­ly 5 and won the Ch­agua­nas West by-elec­tion on Ju­ly 29.This ques­tion at­tempt­ed to probe, in a non-par­ti­san way, whether his in­volve­ment with the Peo­ple's Part­ner­ship Gov­ern­ment for three years be­fore his for­ma­tion of the ILP was some­thing that mat­tered at all to the re­spon­dents in the sur­vey.

There is no doubt that Warn­er is a fac­tor in the minds of peo­ple in the St Joseph con­stituen­cy (86.2 per cent). All oth­er re­spons­es are sta­tis­ti­cal­ly in­signif­i­cant (No-6.3 per cent, Don't know-4.8 per cent and Not sure-2.7 per cent).This con­firms the hy­poth­e­sis that one way or an­oth­er (pos­i­tive or neg­a­tive) Warn­er has be­come a ma­jor fac­tor on the po­lit­i­cal land­scape of the St Joseph con­stituen­cy.

13. Which par­ty can­di­date will you vote for in the by-elec­tion on No­vem­ber 4?

PP–31.2 per cent

ILP–24.6 per cent

PNM–31.4 per cent

MSJ–1.2 per cent

Don't know–7.0 per cent

Not sure–4.3 per cent

Will not say-0.2 per cent

Not vot­ing–0.2 per cent

The ques­tion of vot­er pref­er­ence ahead of the St Joseph by-elec­tion was test­ed in a par­tial­ly hy­po­thet­i­cal en­vi­ron­ment, as on­ly the iden­ti­ty of the PNM can­di­date was known at the time of the sur­vey.There is a sta­tis­ti­cal dead heat be­tween the PNM and the PP, with the PNM frac­tion­al­ly ahead 31.4 per cent to 31.2 per cent.The ILP was the choice of a very sig­nif­i­cant num­ber of re­spon­dents, with a rat­ing of 24.6 per cent.

The un­de­cid­ed vot­ers who re­spond­ed by say­ing "Don't know/Not sure" ac­count­ed for 11.3 per cent, while all oth­er re­spons­es were sta­tis­ti­cal­ly in­signif­i­cant.In as­sess­ing the re­spons­es to this ques­tion, one has to take in­to ac­count the im­pact of the na­tion­al­ly-tele­vised de­bate that took place on the evening be­fore the start of this da­ta-gath­er­ing ex­er­cise and the open­ing of the Grand Bazaar in­ter­change dur­ing the pe­ri­od of the da­ta col­lec­tion.

The nam­ing of the PNM can­di­date ahead of all oth­er par­ties and the on­go­ing con­duct of the lo­cal gov­ern­ment elec­tions cam­paign would al­so have to be con­sid­ered in the as­sess­ment of these re­spons­es.With a mar­gin of er­ror of +/- 4 per cent, it is ap­par­ent that this is a close race.

14. In cast­ing your vote, what will be your main rea­son for your choice?

Po­lit­i­cal leader–12.5 per cent

Can­di­date–9.5 per cent

Par­ty loy­al­ty–16.9 per cent

Need for change–40.0 per cent

Need to re­tain the Gov­ern­ment–17.3 per cent

Don't know–1.7 per cent

Not sure–1.8 per cent

In ex­am­in­ing the choic­es of the re­spon­dents, their re­spons­es in re­spect of their mo­ti­va­tion were in­ter­ro­gat­ed.The largest re­sponse was the "Need for change" which ac­count­ed for 40.0 per cent of the sur­vey.The sec­ond largest re­sponse was the "Need to re­tain the Gov­ern­ment" which got a rat­ing of 17.3 per cent.Par­ty loy­al­ty (16.9 per cent) was the third high­est rat­ing and the "Po­lit­i­cal leader" (12.5 per cent) was the fourth high­est rat­ing.

The choice of can­di­date earned a 9.5 per cent rat­ing, while "Don't know/Not sure" ac­count­ed for 3.5 per cent com­bined.The high rat­ing giv­en to the need for change (40.0 per cent) might be di­lut­ed in a three-par­ty race be­tween two strong op­pos­ing par­ties un­der the first-past-the-post sys­tem of elec­tion.Giv­en the fact that the sec­ond high­est re­sponse "Need to re­tain the Gov­ern­ment" (17.3 per cent) can on­ly ap­ply to the PP Gov­ern­ment, it would ap­pear that this might be­come com­pet­i­tive in a three-way con­test.

15. Peo­ple choos­ing the Peo­ple's Part­ner­ship clas­si­fied by eth­nic­i­ty

Afro–3.0 per cent

In­do–21.8 per cent

Mixed–10.6 per cent

Syr­i­an/Lebanese–0.3 per cent

16. Peo­ple choos­ing the ILP clas­si­fied by eth­nic­i­ty

Afro–8.0 per cent

In­do–10.4 per cent

Mixed–9.5 per cent

17. Peo­ple choos­ing the PNM clas­si­fied by eth­nic­i­ty

Afro–21.9 per cent

In­do–2.6 per cent

Mixed–0.9 per cent

18. Peo­ple choos­ing the MSJ clas­si­fied by eth­nic­i­ty

Afro–0.2 per cent

Mixed–0.8 per cent

In analysing the pref­er­ences of re­spon­dents by eth­nic­i­ty in this mul­ti-eth­nic con­stituen­cy, it is ap­par­ent that the ILP is mak­ing al­most equal in­roads in­to tra­di­tion­al UNC and PNM vot­er blocs.Tra­di­tion­al­ly, the UNC (rep­re­sent­ing the PP here) has been able to win a large ma­jor­i­ty of In­do-Trinida­di­an vot­ers. Tra­di­tion­al­ly, the PNM has been able to win a ma­jor­i­ty of Afro-Trinida­di­an vot­ers.Mixed-race vot­ers have not had a clear pref­er­ence ra­tio when com­pared to Afro- and In­do-vot­ers.

The pres­ence of the ILP has caused al­most even­ly bal­anced re­spons­es across eth­nic­i­ties in their choice of this par­ty when cross-tab­u­la­tions were un­der­tak­en for the 525 ques­tion­naires of the 602 that were ad­min­is­tered. The split be­tween the Afro-vot­ers' pref­er­ence for the PNM (21.9 per cent) and the In­do-vot­ers' pref­er­ence for the PP (21.8 per cent) vir­tu­al­ly off­set each oth­er.The split be­tween the PNM and the PP among mixed-race vot­ers was PNM (10.9 per cent) and the PP (10.6 per cent).

In such a sit­u­a­tion of vir­tu­al even weight­ings be­tween two com­peti­tors, the im­pact of the third par­ty be­comes vi­tal­ly im­por­tant.The ILP's rat­ing with Afro-re­spon­dents (8.0 per cent), with In­do-re­spon­dents (10.4 per cent), and with Mixed-race re­spon­dents (9.5 per cent) would sug­gest that the oth­er re­spons­es that were sta­tis­ti­cal­ly in­signif­i­cant, rel­a­tive­ly speak­ing, could earn a high­er val­ue based on the like­ly close­ness of the con­test.

Both the PNM and the PP ap­pear to have lost sup­port to the ILP in al­most equal num­bers across the board.As a con­se­quence, the out­come of the by-elec­tion re­mains too close to call.