The third race

October 26, 2009 by Ian Ramjohn

In the Sunday Express, Selwyn Ryan wrote

...Mr Panday is also correct when he notes that corruption, on its own, does not decide elections in Trinidad: Race does...

Taran questioned whether this assumption about Trinidadian politics remains valid, pointing out that

If COP support is an indicator, they don't follow their parents. Sure, they didn't win a seat - but they did cast a wide net on Election Day because something about COP was appealing. It wasn't Winston Dookeran's charismatic speeches. What was it?

If you're interested in predicting outcomes, phenomonological models are good enough.  Race, coupled with a few other factors, can explain the outcome of every election since 1956 with the possible exception of 1986.  If you consider mixed people as a "swing" constituency, caught between the two races, you've got your model.  Figure out how they'll vote/how they voted, and you've predicted or explained your election.  It works, it's simple...it's what they call an elegant solution.

The problem with phenomonological models is that they are only interested in emergent patterns, not in the mechanisms that generate those patterns.  Correlation, they tell you, does not imply causation.  If you want to study the underlying mechanics that create those patterns you need to dig deeper, and look for more complicated models.

To understand what's really going on, you need to look at Trinidadian society at a somewhat finer grain.  Lloyd Best's subdivision, with its Afro-Saxons, Garveyites, Hindus and Presbyterians, may be a bit too fine-grained.  While these groups exist (or existed), I don't think they are a useful proxy to explain voting behaviour.  Neither NJAC nor the Liberal Party ever attracted more than a tiny sliver of the electorate.  Nor did the Butlerites post-1956.  But four parties have managed to attract more than a sliver of the popular vote - POPPG, ONR, the rump-NAR of 1991 and COP.  Throughout the history of party politics in Trinidad, there have been three key political constituencies.  The race-based model only works when the third constituency is disorganised.  (More to come.)

Permalink