Aubrey Stewart | Beyond PNP or JLP country: Jamaica’s performance voters

For generations, Jamaica’s political landscape has been drawn as “PNP country” or “JLP country”. Here, party loyalty has often been more than a choice made on Election Day. It can feel almost inherited, passed down like genes from one generation to the next, reinforced by community, and woven into political identity.

But, what happens when neither party can automatically claim your vote? When the political identity you inherited no longer tells you who to vote for?

Perhaps the most interesting lesson from the 2025 general election was what happened to those who neither committed to JLP nor PNP.

My recent study, Beyond Party Loyalty, examined a national pre-election survey conducted between May and August 2025, linking how Jamaicans assessed political performance with official election results.

There is an obvious difficulty in asking voters whether government performance shapes their choices. A JLP supporter may genuinely believe a JLP government is performing well. A PNP supporter may look at identical conditions and reach the opposite conclusion. Political science describes this partly through ‘motivated reasoning’: our political identities influence how we interpret information. So, when someone says the government is doing a good job and that they intend to vote for it, we cannot conclude that performance caused the decision. Existing loyalty may have produced both answers.

But this gives us a way to test the question rather than argue about it. If performance ratings are mostly loyalty in disguise, the link between performance and vote should be strongest among committed partisans, who have the clearest motive to talk themselves into their own side’s record, and weakest among those with no party at all.

The study separated respondents into JLP supporters, PNP supporters and those declaring no affiliation. Of the 6,334 respondents, 48.3 per cent identified with the JLP and 40.8 per cent with the PNP. Only 10.5 per cent, or 663 people, were non-partisan. That small group produced the central finding.

WHEN LOYALTY IS WEAK

Among JLP supporters, a favourable evaluation of government performance was associated with a 14.6 percentage point increase in the probability of intending to vote JLP. Even among PNP supporters, it increased the probability of intending to vote JLP by 16.8 points. Among non-partisan Jamaicans, however, the increase was 41.2 percentage points, nearly three times as large.

The pattern held for evaluations of the Opposition. Among non-partisans, a positive assessment of the opposition was associated with a 29.7-point reduction in support for the incumbent, roughly twice the effect found among either party’s supporters. An analysis restricted to the 663 non-partisans produced almost identical results.

Performance mattered most where party loyalty was weakest.

While 48.4 per cent rated the government’s performance as good or very good, only 27.6 per cent said the same of the Opposition. The unaffiliated voter was asking whether a credible alternative existed, and fewer than three in ten thought so.

Economic management had the strongest association of any policy area: 22.1 points among non-partisans, against 3.1 among JLP supporters and 7.9 among PNP supporters. Health and infrastructure followed at 13.3 and 10.7 points among non-partisans, with no measurable effect among committed supporters.

Security and education showed no independent relationship among non-partisans once other factors were accounted for. That may be because judgements about crime and schools were absorbed into the broad verdict on competence rather than counted separately, or simply because 663 respondents is a thin base, once divided across five policy areas. Either way, it cautions against assuming that every government achievement produces an electoral reward.

LOCAL LEADERSHIP STILL MATTERS

Voters also evaluate the people representing them, and the same structure appears. Among JLP supporters, a favourable MP rating was associated with a 7.3 point increase in support for that MP’s party. Among PNP supporters, the figure was also 7.3 points, statistically indistinguishable between the two groups. Among non-partisans, it was 28.5 points, roughly four times as large.

A popular party brand cannot permanently substitute for an effective representative. But the reverse deserves saying too. Among the 88 electoral divisions within Jamaica’s historically competitive constituencies, roughly one in five was won by an MP whose own constituents rated him poorly. Partisan strength still rescues weak representatives. It simply no longer does so reliably.

All of this rests on what people told an enumerator. The harder test is whether their assessments track what returning officers certified.

For two measures they do. Aggregating government performance ratings to the electoral division and matching them against official returns, moving a division from the bottom quarter of ratings to the top quarter raises the predicted probability of an incumbent victory from roughly 5 per cent to roughly 93 per cent. MP performance survives the same test.

Opposition ratings do not survive aggregation, and neither does any individual policy domain. That is a real limit, and the policy-specific results should not be carried across into claims about who wins seats.

NO LONGER A TRIBAL QUESTION

In a country where fewer than four in 10 registered electors now vote, one result deserves particular attention. Party identification is a powerful predictor of how a Jamaican votes. It is not a predictor of whether a division turns out at all. Identifying as a JLP or PNP supporter had no meaningful association with divisional turnout. Performance ratings did, with a favourable government rating linked to a 3.3-point increase in the probability of a division recording high turnout.

The traditional assumption was that the machine delivered bodies to the polling station. On this evidence, what brings people out now looks more like a verdict on delivery than a summons from a party.

None of this means Jamaican politics has become non-partisan. Nearly nine in ten respondents identified with one of the two parties, and partisan identification remained among the strongest predictors of vote intention. A caution is owed as well. This study cannot establish that performance evaluations cause voting decisions. Its cross-sectional design means the findings are properly understood as associations, and the original paper is explicit about that.

Still, the pattern is difficult to ignore. It replicates across three separate performance measures, holds when the unaffiliated are analysed on their own, and survives the move from what people said to what the returns recorded. Elections are not decided by the largest group. They are decided at the margins, and the 10 per cent without a declared affiliation carry influence out of proportion to their numbers in competitive seats.

There is an encouraging lesson in that. Jamaica’s history of partisan loyalty does not prevent electoral accountability. The two coexist. Committed supporters read politics through inherited identities, while another segment judges governments, policies and representatives conditionally. Parties will always need loyal supporters. But the voters who determine who governs increasingly ask a simpler question: what have you actually delivered?

Dr Aubrey Stewart is a public policy researcher and consultant. Send feedback to aubreymstewartiv@gmail.com