Forecasting the 2016 General Election in Jamaica

Abstract

The objective of the article is to explain the methodologies and the findings of the 2016 Jamaican General Election forecasts. The Good Judgment Project’s CHAMPSKNOW system was applied using qualitative and quantitative methods. The research question was: what were the probabilities of the Jamaica Labour Party (JLP) or the People’s National Party (PNP) winning the 25 February 2016 General Election? The data were drawn from election results and macro-economic variables from 1962 to 2015; polls from 1976 to 2016; campaign ads; election newspaper stories; constituency fund disbursements and independent surveys in marginal seats. The results showed that the JLP had a greater number of ads with better and clearer policy contents than the PNP ads. The JLP also received more positive news coverage during the campaign. MPs who spent a large part of their constituency funds on welfare were more likely to win. The PNP had more garrison, traditional and marginal seats than the JLP so the PNP had the edge. Moreover, the data from the independent surveys and the macro-economic analyses indicated the likelihood of a PNP win. The national polls revealed a statistical dead heat but the forecasts started with the governing PNP having a slightly greater probability of winning because of its active political business cycle in which the PNP was rolling out programmes throughout the country in the election year. The forecasts were revised when the JLP narrowed the gap because the PNP refused to participate in the national debate, which generated negative news about the PNP. The final forecast said the election would be close with the PNP having the edge. However, the JLP’s tax plan was a wild card, which gave the party the edge with a one-seat victory.

Keywords:

• political forecasting
• General Election
• Jamaica
• CHAMPSKNOW

Acknowledgements

This paper is dedicated to the late Louis G. Lindsay, retired lecturer of the University of the West Indies, Mona who transitioned on 25 February 2016. We would like to thank statistician Dr Novie Younger-Coleman and econometrician Dr Nadine McCloud for their advice. However, the authors take full responsibility for the shortcomings of the final forecast and this article.

Disclosure statement

No potential conflict of interest was reported by the authors.