Grenada Election 2026: NDC's Seat Map vs. the Polls as Voters Choose a Prime Minister

Grenadians vote today, 5 November, for all 15 seats in the House of Representatives, and the two best guides to the result disagree with each other in a way that tells you something about how elections actually work. A face-to-face poll says voters prefer the opposition leader; a constituency-by-constituency model says the government will probably keep its majority anyway. Both can be right about what they measure. Only one of them will be right about who forms the next government.

The poll comes from NACTA, the North American Caribbean Teachers group, which interviewed 1,285 voters in person across 14 constituencies, leaving out Carriacou, with a margin of error of four percentage points. It puts opposition leader Emmalin Pierre of the New National Party (NNP) ahead as preferred prime minister, at 47%, four points clear of incumbent Dickon Mitchell of the National Democratic Congress (NDC), with Peter David of the newly formed Democratic People's Movement (DPM) at 4% and 6% undecided or offering no choice. The poll's authors describe an electorate leaning toward change: dissatisfaction with the cost of living, fuel prices, unemployment and healthcare affordability, grievances about governance and alleged political victimisation, and anxiety over United States visa restrictions that have made a foreign-policy question feel personal, affecting artistes trying to tour, families trying to visit relatives, and businesses trying to travel. Pierre's performance in the leaders' debate appears to have helped her, and the prospect of Grenada electing its first woman prime minister is plainly part of her appeal.

The model, built by analyst Michael Derek Roberts from the 2022 results, constituency strengths and weaknesses, and thousands of simulations, gives Mitchell's NDC a 76 to 81% chance of winning the eight seats needed to govern. Its single most likely outcome is NDC 10, NNP 5, with David's DPM shut out, though the average across all simulations is closer to nine seats for the NDC — a gap that means several of those ten seats are far from secure. The NDC's path runs through six safe and likely seats plus any two of four battlegrounds: Carriacou and Petite Martinique, where Tevin Andrews is favoured at 67 to 72%; the Town of St George, a genuine three-way toss-up where David's personal following splits the opposition vote; St Andrew North West; and St Patrick East, the seat to watch.

I find the tension between these two readings more informative than either one alone, because they measure different things. A national preference question is not a seat count. Grenada's 15 constituencies are won one at a time, and a mood for change spread thinly across the country can coexist with a government majority built on loyal voters concentrated in the right places. That is the NDC's real advantage: the poll itself notes that its supporters remain strongly attached to their party, while the opposition's hopes depend on two things that are harder to guarantee — that NNP supporters turn out in large numbers, and that undecided and unaffiliated voters break decisively toward change. The DPM complicates that further. David is not contesting enough seats to govern, but in tight constituencies his votes could decide whether the NNP or the NDC wins the seat, and in the Town of St George that split currently helps the government.

None of this makes the NDC's position comfortable. A 76 to 81% model probability built on 2022 results is a statement about the map, not about today's mood, and the freshest evidence we have — the in-person poll, the debate, the economic grievances — all points the other way. If NNP turnout surges and the undecideds break for change, the model will look like a relic of a calmer election. That is precisely what makes this race worth watching: it is a test of whether a constituency map drawn from the last election can survive a change election, or whether a national mood can overpower geography. Polls in Grenada open at 6 a.m. and close at 5 p.m. local time, and the count will answer a question that applies well beyond one small island — when the polls and the map disagree, which one do you trust?

My judgment is that the map holds, narrowly. I put the NDC's chances of winning at least eight seats at 65%: below the model's range, because the poll's change appetite and Pierre's debate momentum are real and recent, but still the likelier outcome, because the NDC starts with a loyal base, needs only two of four battlegrounds, and benefits wherever the DPM splits the anti-government vote. If I am wrong, the interesting error will not be that a poll beat a model; it will be that turnout, the one thing neither instrument measures well in advance, decided the government.