1 / 22
WHY JAMAICANS VOTE:
IMPLICATIONS FOR WHO WINS THE 2016 GENERAL ELECTIONS
Herbert Gayle, PhD (London), MSc, BA (UWI)
Methodological Note
This study is a single snapshot of the 63 constituencies across Jamaica; and utilizes integrated
methodologyas the main approach, with the assistance of qualitative and historical material. In simple
terms, integrated methodology (started in the 1930s but developed and formalized in the 1990s) is the
meticulous combination of qualitative and quantitative research methods to fit specific philosophical
objectives and research tasks. In other words, it is not simply about mixing methods, but using specific
tools and dovetailing them in specific frames to meet the objective of a particular study. For this study a
4‐question standardized mini‐integrated instrument (lasting no more than 3 minutes) was constructed
utilizing open‐ended questions, which allowed persons to freely express their views and positions.
Responses were analyzed using 3 categorical variables: gender, age, and SES. Socio‐economic status was
determined by combining data on area of residence, parents and/or respondents’ occupation, and level
of education. While there are clear distinctions between all SES groups, we conveniently combined
middle and upper middle income groups (UM) and poor and near‐poor (PN). These combinations are
well known to work in market and political research efforts. The former group range from merchants
through upper middle class professions such as law and engineering to the upper bands of the nurture
and security professions; the latter ranges from lower bands of the nurture and security professions
through artisans to unskilled labour. Socio‐economic status was treated as an emergent variable. This
means that the proportion in the sample of each constituency is likely to reflect the economic
demographics of the constituency, given no quota was set – unlike age (18‐34 described as ‘Young’ and
35+ years described as ‘Mature’) and gender/sex(Male/Female)where there was an attempt to sample
equal amounts.
Though this study cannot be treated as a poll, it used 2 geo‐political variables ‘parish’ and ‘constituency’
and hence can provide rough insights into the direction of voting or assist with a seat count. The
research team was trained to cover all the polling divisions in each constituency, paying attention to
proportion, to allow for some degree of accuracy. Nonetheless, the study was not designed in a way to
provide precise data such as X will win by X% ‐ and this was not the main objective of the study. The
main reason for this inability is the fact that the study was set on a ‘bare essential’ sample.
Consequently, the confidence for seat counting is conservative. In the findings we use categories such
as:
1