POLL­STER AND PRO­FES­SOR HAMID GHANY

The poll was con­duct­ed from Thurs­day, April 10, to Sun­day, April 13, 2025. A to­tal of 1,650 ques­tion­naires were ad­min­is­tered in Trinidad over 11 con­stituen­cies that were se­lect­ed based on a mar­gin­al­i­ty for­mu­la that I in­vent­ed in 2002. The mar­gin of er­ror for the Trinidad poll is +/- 2.5 per cent.

A to­tal of 530 ques­tion­naires were ad­min­is­tered in To­ba­go—265 ques­tion­naires in To­ba­go East and 265 ques­tion­naires in To­ba­go West. The mar­gin of er­ror for the To­ba­go poll is +/- 4.0 per cent.

Cal­cu­la­tion of mar­gin­al­i­ty in Trinidad

In con­duct­ing the na­tion­al sur­vey, it was de­cid­ed to use the par­lia­men­tary elec­toral con­stituen­cies as a nat­ur­al ge­o­graph­i­cal ba­sis for gath­er­ing the da­ta. To this end, the of­fi­cial re­sults of the gen­er­al elec­tions of Au­gust 10, 2020, pro­vid­ed the ba­sis for the choice of con­stituen­cies to­geth­er with the bound­ary and con­stituen­cy changes con­tained in the Re­port of the Elec­tions and Bound­aries Com­mis­sion dat­ed March 13, 2024, to­geth­er with my mar­gin­al­i­ty for­mu­la.

A cal­cu­la­tion of mar­gin­al­i­ty (M) was made to con­firm what were the pro­ject­ed mar­gin­al con­stituen­cies for the 2025 gen­er­al elec­tion. The sym­bols for this for­mu­la are M for mar­gin­al­i­ty, and D for the mid­point 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 nor­mal­ly save their de­posits be­cause 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

As a con­se­quence, the fol­low­ing con­stituen­cies in Trinidad can be con­sid­ered mar­gin­al in re­la­tion to the 2025 gen­er­al elec­tion, ow­ing to the fact that their pro­ject­ed mar­gin­al­i­ty (cal­cu­lat­ed dif­fer­ence mea­sured be­tween the first and the sec­ond can­di­dates based on 2020 elec­toral da­ta and 2024 bound­ary changes) fell in­side of > 3,504:

1. Aranguez/St Joseph (823)

2. Barataria / San Juan (1,060)

3. Moru­ga/Table­land (1,072)

4. Ch­agua­nas East (1,086)

5. San Fer­nan­do West (1,618)

6. Clax­ton Bay (1,813)

7. Tu­na­puna (1,927)

8. La Hor­quet­ta/Tal­paro (1,939)

9. To­co/San­gre Grande (2,976)

10. Cu­mu­to/Man­zanil­la (3,326)

11. Ma­yaro (3,364)

These con­stituen­cies were polled by us­ing three PDs per con­stituen­cy on the ba­sis of choos­ing one very strong PNM PD from 2020, one very strong UNC PD from 2020 and one mar­gin­al PD from 2020 at 50 ques­tion­naires per PD.

ANALY­SIS

Po­lit­i­cal par­ty sup­port

The per­for­mance of the PNM Gov­ern­ment over the last ten years, which is es­sen­tial­ly con­nect­ed to the lega­cy of the for­mer Prime Min­is­ter, Dr Kei­th Row­ley, has placed the new Prime Min­is­ter, Stu­art Young, at a dis­ad­van­tage, as he has to ei­ther ac­knowl­edge that he was part of this per­for­mance or he is go­ing to dis­tance him­self from it.

The Leader of the Op­po­si­tion, Kam­la Per­sad-Bisses­sar, has re­brand­ed her­self as the leader of a new coali­tion of in­ter­ests which has broad­ened the base of the UNC by bring­ing per­son­al­i­ties who were pre­vi­ous­ly op­posed to her in­to her fold just as they have changed their dis­po­si­tion to­wards her.

The opin­ion poll re­spon­dents have tak­en a some­what pos­i­tive view (45 per cent) of the UNC coali­tion of in­ter­ests, while the PNM has not been as favourably viewed by re­spon­dents in the poll (30 per cent).

The split-vote phe­nom­e­non is alive and well, as the com­bined ef­fect of the sup­port for the Pa­tri­ot­ic Front (7 per cent) and the Na­tion­al Trans­for­ma­tion Al­liance (6 per cent) has a 13 per cent vot­er de­vi­a­tion loom­ing over the heads of both the UNC and the PNM.

As re­gards which of the two ma­jor par­ties they are hurt­ing more, it is dif­fi­cult to tell. If it is the UNC Coali­tion, then that would mean that they would have to make fur­ther in­roads in­to the UNC Coali­tion to hurt their chances. If, how­ev­er, they are hurt­ing the PNM as part of an an­ti-UNC vote, then the UNC will be the prime ben­e­fi­cia­ry of such a vote split.

This out­come was gath­ered from 11 mar­gin­al con­stituen­cies, and a 13 per cent vote split will be sig­nif­i­cant un­der the first-past-the-post sys­tem of elec­tion. If the PF and the NTA are cut­ting in­to UNC coali­tion votes, then more cut­ting will be re­quired to re­duce the chances of the UNC coali­tion. If the PF and the NTA are cut­ting in­to PNM votes, then that does not au­gur well for the PNM.

How the ac­tu­al vote split­ting will play out in any of these mar­gin­al con­stituen­cies can­not be pre­dict­ed, but any close fin­ish­es could be the re­sult of vote split­ting one way or an­oth­er.

This poll was tak­en two weeks be­fore the Gen­er­al Elec­tion and so the snap­shot may change if there are any switch­es among third par­ty votes as a re­sult of the fi­nal two weeks of cam­paign­ing.

Be­liev­abil­i­ty/Good for the Econ­o­my Ques­tions

The Be­liev­abil­i­ty/Good for the Econ­o­my ques­tions showed that the Leader of the Op­po­si­tion en­joyed a lead over the new Prime Min­is­ter that was con­sis­tent with the choice of par­ty in the fi­nal ques­tion of the sur­vey. The sep­a­ra­tion of these ques­tions be­tween the start of the poll and the end of the poll was de­signed to see whether there would be a mod­icum of con­sis­ten­cy be­tween these first re­spons­es and the last re­sponse.

Es­sen­tial­ly, there is a cor­re­la­tion be­tween them. How­ev­er, in a first-past-the-post sys­tem, one does not know how the fi­nal re­sults will play out in in­di­vid­ual mar­gin­al con­stituen­cies.

The re­spons­es to these ques­tions may be in­dica­tive of re­spon­dents seek­ing a change now that Dr Row­ley is gone, and they may not be re­spond­ing as pos­i­tive­ly to Prime Min­is­ter Young as his re­place­ment.

On the flip side, the Leader of the Op­po­si­tion, Kam­la Per­sad-Bisses­sar, has formed new al­liances which may have been viewed pos­i­tive­ly by the re­spon­dents as a new im­age that they may be will­ing to try.

TO­BA­GO

The two To­ba­go con­stituen­cies were polled us­ing five PDs in each con­stituen­cy based on trend­ing re­sults us­ing 2020 gen­er­al elec­tion da­ta and track­ing those PDs through the Jan­u­ary 2021 and De­cem­ber 2021 To­ba­go House of As­sem­bly (THA) re­sults.

The track­ing of these re­sults for the ini­tial PDs from 2020 through the two THA elec­tions of 2021 was deemed to be nec­es­sary see­ing that (i) the PNM lost ground, (ii) the PDP gained ground, and (iii) the TPP was sub­se­quent­ly formed out of the PDP. Ge­o­graph­i­cal spread was al­so a fac­tor in se­lect­ing PDs.

The method­ol­o­gy had to con­tend with the ef­fect of the de­cline of the PNM in three elec­tions (2020 Gen­er­al Elec­tion, Jan­u­ary 2021 THA elec­tions and De­cem­ber 2021 THA elec­tions) and whether they would re­gain any lost sup­port be­cause of the split be­tween the PDP and the TPP. At the same time, there was the need to mea­sure whether the PDP would hold its 2021 vote lev­els or whether the TPP had in­her­it­ed the votes of the PDP af­ter the split be­tween them in 2022/2023.

ANALY­SIS

Po­lit­i­cal Par­ty Sup­port

The like­li­hood of split vot­ing is very ap­par­ent in To­ba­go giv­en the per­cent­age dif­fer­ences be­tween the three main par­ties in the con­test on the is­land. The PNM is lead­ing with 47 per cent; how­ev­er, the TPP, which was formed out of the PDP, is at 32 per cent and the PDP is at 9 per cent. There was al­so 4 per cent of re­spon­dents who in­sist­ed to the field da­ta-gath­er­ers that they were keep­ing their choic­es pri­vate and did not want to share that in­for­ma­tion. Al­so, 3 per cent of the re­spon­dents did not know. That cu­mu­la­tive 7 per cent re­mains an un­known fac­tor, while 4 per cent of the sur­vey said that they would be vot­ing for oth­er par­ties/can­di­dates.

In a tight race where up to 20 per cent of the sur­vey is not clear­ly iden­ti­fied with ei­ther of the main par­ties (PNM or TPP), the two seats in To­ba­go could go the way of the PNM if the vote-split­ting con­tin­ues to trend in favour of the PNM. On­ly one per cent said that they would not sup­port any of the par­ties con­test­ing the race.

As re­gards which of the two ma­jor par­ties is hurt­ing more as a con­se­quence of this split-vote phe­nom­e­non, it ap­pears that the TPP is more like­ly to be af­fect­ed by the loss of votes to the PDP since it is the par­ty out of which the TPP emerged.

The Be­liev­abil­i­ty Ques­tion

The Be­liev­abil­i­ty ques­tion showed that the new Prime Min­is­ter en­joyed a lead over the Chief Sec­re­tary of the THA which was con­sis­tent with the re­sults for the PNM on the ques­tion of choice of par­ty.

How­ev­er, the Chief Sec­re­tary (36 per cent) scored high­er than his par­ty (the TPP) did in the ques­tion of choice of par­ty, which was 32 per cent. In­ter­est­ing­ly, 8 per cent said that they did not know, 9 per cent said nei­ther of them and 1 per cent said Wat­son Duke.

The re­spons­es to these ques­tions may be in­dica­tive of re­spon­dents seek­ing a change now that Dr Row­ley is gone, and they may be open to giv­ing the new Prime Min­is­ter, Stu­art Young, a chance as his re­place­ment.

On the flip side, the Chief Sec­re­tary has sur­vived the split in the PDP, and his ad­min­is­tra­tion in the THA has re­mained in­tact dur­ing the tran­si­tion from be­ing in­de­pen­dent mem­bers of the THA to be­com­ing mem­bers of the TPP that was formed in 2023. There has been a lot of con­tro­ver­sial com­men­tary be­tween him­self and the for­mer prime min­is­ter as well as the leader of his for­mer par­ty, the PDP. Such con­tro­ver­sial com­men­tary does take a toll on pub­lic imag­ing, and, al­though he is not on the bal­lot, his par­ty—the TPP—cer­tain­ly is.

Con­clu­sion

This gen­er­al elec­tion is like­ly to be af­fect­ed in mar­gin­al con­stituen­cies in Trinidad and the two To­ba­go con­stituen­cies by split vot­ing in­volv­ing third par­ties. In any close con­tests in these con­stituen­cies the pen­du­lum could swing one way or the oth­er be­cause the quan­tum for third par­ties, oth­ers, not sure and pri­vate views takes the fig­ures in­to dou­ble dig­its.

Any late swing in the fi­nal two weeks of the cam­paign af­ter this sur­vey was done could ei­ther ex­ac­er­bate the vote split or re­duce its im­pact. So far, it ap­pears to be a fac­tor.