Robert Buddan, Contributor

We have little proper knowledge of small democracies, especially mini and micro-democracies. We tend to generalise, say, about voting behaviour, the role of government, or the impact of personalities across democracies even though we should know that one size does not fit all. We also do this when we study economies and social behaviour.

So, businesspersons used to one environment might make bad investments in another, or migrants might not settle well into another country's social structure. IMF policies for some kinds of countries have failed when not adjusted for other kinds of countries.

We have some information about small economies in international trade and small ecological systems facing global climate changes. But political studies lag behind when it comes to understanding the ways in which small political systems might behave differently than larger ones. We are then caught by surprise when an election produces a result that goes radically against what standard polls had predicted and what conventional wisdom had expected.

St. Lucia can be described as a mini-democracy. The island covers 234 square miles, has a population of 168,000, an electorate of 135,000, and 17 constituencies. It can be difficult to predict elections in such small democracies. The reason is simple. It takes changes in a small number of votes in a small number of constituencies to make the all-important difference as to which party wins and which loses. Those small numbers of votes can even produce exaggerated seat differences, making the outcome seem greater than what had been predicted.

Size and predictability

Much is being made of two things - that incumbents can lose and that polls can be wrong. Both might be even truer in small democracies. Small size could make it difficult to track changes in opinion, which can occur more suddenly and immediately because small space transmits impact events more immediately and more intimately.

A Barbadian polling organisation had suggested that the St. Lucia Labour Party could win 10 seats to seven for the United Workers Party. The UWP won 11 to six for the SLP instead. This is a difference of four seats. Apparently, it was also a difference of 2,000 votes. Many seats were closely won by the UWP and closely lost by the SLP. The difference of 2,000 votes would have been within the margin of error of the Barbadian poll. It is not at all astonishing for this kind of margin to change in a week or two.

Voter turnout also has a telling effect, especially in small societies, where even a small difference between those expected to vote for a party but did not vote at all can be significant to the outcome. The Bill Johnson poll had predicted a wide margin of victory for the SLP. But his final poll had suggested that only 10 per cent of respondents would not vote. In the end, 40 per cent did not.

Worse, if SLP supporters felt too confident by the polls, many more of them might have stayed home relative to UWP supporters. The problem with the two polls might simply have been that they did not track likely voters and were not done late enough to track late changes in voters' decisions.

This is important because of other differences about small democracies. The small space that the population shares means that there is hardly any difference between the local and the national. Something of local impact is almost always also of national impact. St. Lucia has a crime problem. A murder in one locality will trigger anxieties across the nation easily. St. Lucia's major industry, bananas, has suffered from new European trading rules. The impact is not merely rural. It is national.

The country's only important industry is now tourism. Crime in the tourism area will have a magnified effect because of small size and the disproportionate importance of tourism to the population relative to the country's economy. It appears that two issues - crime and the decline of the banana industry - hurt the SLP the most. But these also have to be seen in the context of the small society and economy. These events did impact after the last polls were published.

Dominant personalities

Small societies also produce a narrower field of leading personalities. These societies are more easily dominated by a few strong or well-known personalities. Those personalities can more easily command the resources of politics or economy to dominate public life. Such personalities can have strong late impact in these societies.

Small democracies have produced such personalities like Vere Bird of Antigua, Eric Gairy of Grenada, Robert Bradshaw of St. Kitts, Eric Williams of Trinidad and John Compton of St. Lucia. In many cases only illness or death was able to get the better of them.

Sir John Compton is St. Lucia's historic political personality. Not even the regionally known Vaughan Lewis, nephew of St. Lucia's Nobel laureate, Sir Arthur Lewis, could command anything like the loyalty Sir John could when Lewis took over the UWP after Compton retired.

Compton was head of government for 29 years before and after St. Lucia's independence. He led the country to independence and is regarded as Father of the Nation. He returned to politics in 2005. At the age of 82, Compton remains a towering figure in the country.

Even so, Compton was not a sure thing to win and even a week before the elections seemed to be a likely loser. But he obviously knew something about the politics of small size. He promised St. Lucians a safe and secure society reminiscent of the past, capitalising on the anxieties over crime and the banana crisis, while using the deep resonance his stature enjoyed among the people.

Compton learned something about the national psychology of his people over his long political life. Small societies tend to feel particularly vulnerable. His opponents accuse him of a dirty campaign of lies about government corruption. But this was designed to feed into his 'trust me' campaign. It was well orchestrated for that kind of society.

Already some opposition leaders are taking heart from Compton's win, especially ageing leaders like Lester Bird of Antigua. Parties that are behind in polls are now happy to attack pollsters. Compton's win does not necessarily signal a trend for opposition parties. One event does not indicate a trend. Each election is best seen in its own time and place. While there are always lessons to learn from elections it might be that the lessons are best transferred to comparable cases like other small democracies in the Eastern Caribbean.

In the final analysis, a sound polling technique cannot be divorced from solid political sociology. It is important to know the history of an electorate and its voting behaviour, and the history of constituencies, parties and candidates to understand the nuances of a society. This will help to indicate how stable and deeply held the opinions of respondents are, or the extent to which their responses can be trusted.

We should remember that Carl Stone was not simply a pollster. He was a political sociologist. Polling techniques are important but one method might not fit all societies equally well because all societies do not fit one theory.

Robert Buddan lectures in the Department of Government, Mona, UWI. Email: Robert.Buddan@uwimona.edu.jm.