PETER WICKHAM: Reflections on Polling II
There are several analytical perspectives that arise from the outcome of the 2015 United Kingdom general election. however, the one that is most intriguing to the pollster is the undeniable fact that the polls there failed to predict the actual outcome of the election, and this naturally raises questions regarding the utility of polling.
This is an issue which I am anxious to engage for reasons which are not entirely selfless, especially as similar concerns have been raised in Barbados recently (although the two scenarios are not entirely analogous).
This type of outcome is always the possibility that any pollster fears. however, the reality of the Caribbean environment is such that this type of polling result can impact more negatively on the pollster’s reputation and livelihood.
It is therefore fortuitous that the events of last week occurred in the UK for several reasons which are worthy of mention. Most important is the fact that the UK is an entirely more mature political environment, where the science of polling is accepted and trusted.
In the Caribbean, polling is relatively more novel and the reality of our propensity towards “yardfowlism” is such that a poll which runs counter to the wishes of a political party is likely to be dismissed as partisan propaganda.
Recently our leader suggested that he “ignores” polls, which is an honest reflection of the contempt other regional leaders hold more silently. In the UK, however, no leader would dare dismiss poll data in similar fashion unless he had statistical evidence to the contrary, which is reflective of an admirable level of political maturity.
Without digressing too much, it should be a noted that a similar level of maturity is likely to have influenced the rejection of Miliband’s Labour platform which was in many ways as unbelievable as the Democratic Labour Party’s promise not to send home a single worker or divest any state enterprise.
Certainly the nakedness of Miliband’s central manifesto promises was exposed shortly after its release when a rigorous process of “fact-checking” showed the extent to which he appeared not to have properly considered the cost of his vote catching initiatives, or the potential impact of the UK budget deficit.
While this was only partially true, the UK electorate has clearly moved beyond the mental separation of electioneering from governance, which is one of the hallmarks of a mature democracy.
Another reason why it is fortunate that this polling debacle occurred in the UK is related to the large number of polls that are conducted there leading up to any election. Here in the Caribbean we consider ourselves fortunate to have two competing polls, while in the UK there were easily half-dozen in the days before the election.
The fact that so many polls commissioned by different agencies all said the same thing would make it impossible to suggest that the “error”
is associated with the bias or incompetence of any single pollster.
The number and frequency of polls which ran up to the day before the election also help to place in an appropriate context the suggestion that a “second poll” conducted a week after the first was in any way irregular.
Having established the foregoing, the next logical focus is the factors motivating people to either lie or change their mind and in this regard my comments are entirely speculative since I can claim no superior knowledge of the British elector’s mind.
The Barbados experience is helpful since electors here openly discussed the factors that motivated them to vote in a way that was different to the general direction of the polls.
We can eliminate vote buying as a factor there since that practice tends not to work well in large mature democracies. Instead, several voters speculated about their familiarity with the polls which motivated many of them to act in defence of democracy, especially as the Grenada election presented an example (two-days before) which would “spook” any right thinking guardian of democracy.
I am inclined to think that the British voter was similarly minded and voted against the background of knowledge of poll projections with which they were uncomfortable.
The polls suggested a hung Parliament with the possibility of a Labour-SNP coalition led by Ed Miliband in whom the British voter did not repose much confidence.
This was to many a toxic mix which could easily have returned the vexed issue of Scottish nationalism to the fore, especially since Miliband’s political savvy could not be relied upon to avoid such uncomfortable predicaments.
The British voter therefore appeared to have voted strategically to avoid any such possibility, by voting more heavily in favour of a leader they were familiar with and who performed at least well enough to be trusted again.
It could therefore be argued that voters who responded to pollsters were honest regarding their intentions up to the day of the elections, but these intentions were modified based on their consideration of the likely impact of the voting intentions of others.