By Michael Marsh, Emeritus Professor, Trinity College Dublin

Election analysts have distinguished the long campaign from the official campaign and while the latter here lasts just three to four weeks, the former really runs from when the results of the previous election are known and the government is formed.

However, this long campaign can become more intense as the official, short one draws near.

The parties have now been in campaign mode for some time, at least since the Budget last October and arguably for some time before that.

The earliest studies of voter behaviour, using opinion polls which were carried out from the 1940s onwards in the United States, initially assumed that the repeated surveys carried out throughout the last few months of the campaign would reveal to researchers how the people made their choice.

Such studies were new and it might be expected that they would reveal things we did not know before, but one thing in particular surprised them.

This was that most people seemed to have made their decision long before the campaign started.

For many, they had what a later observer called a "standing decision" to support a particular party, and while they might well follow the campaign, their reading of it would be filtered through a partisan screen.

Similar results were found when similar studies began to be carried out in other democracies, and in Ireland the limited survey research that was carried out from the late 1960s confirmed what many political activists on the ground knew only too well: that most had a standing decision to vote for Fianna Fáil or Fine Gael.

This pattern of stability was probably exaggerated: after all, Fine Gael support could fluctuate considerably. Yet this remained a substantial "truth" of Irish politics into the 21st century.

Yet if we look at the exit polls that have been conducted for RTÉ from 1997 onwards, for which respondents are sought as they come of of their polling stations, we see signs that a very substantial number of people do change their vote from election to election.

In the 2007 poll, for instance, we find that more than 20% of those voting FF or FG in 2002 voted differently in 2007, and an even higher percentage of those who voted for someone else changed their preference, with almost 30% in all switching parties.

This finding relies on asking people how they voted five years ago, but memory is selective, and people tend to misremember in ways that makes them seem more consistent.

The survey data gathered for the Irish National Election Study 2002-07 allows us to carry out a similar analysis but this study interviewed the same people on each occasion.

Results indicated that the real level of switching was almost 10% higher, closer to 40%. And yet the overall results in 2002 and 2007 were remarkably similar.

The exit poll indicated that almost 50% of those voting in 2007 and 2011 switched their allegiance, and we can assume the real figure was even higher: the majority of voters switched.

Current polls are suggesting more change again. A Poll of Polls analysis combining recent RED C and B&A polls – both asked about the 2011 vote – suggests that more than a third of those who voted in 2011 are currently intending to vote differently this time.

Exit polls have asked people when they finally make up their mind which party or independent candidate to vote for. This question suggests that about half of those voting – and more in 2011 – make up their mind in the last three weeks, during the official campaign.

The 2002 and 2007 Election Study surveys, which interviewed people in the weeks after the election,  put the percentage of late deciders much lower, at just under a third.

Possibly, the difference is because of the timing of the different studies; those who have just voted have just implemented a decision, whereas reflecting on it several weeks later a respondent might be more inclined to take a long view. On either account there seems much to play for.

Yet a substantial body of political science research does take the view that such change as does take place in the campaign is not unpredictable. Moreover, to a considerable degree the changes during the campaign tend to bring voters into line either with long-term loyalties, or predispositions.

The long-term loyalties of the Irish voter now seem more than a little problematic.

While more than two thirds of voters in the late 1970s reported that they would see themselves as close to a particular party, this figure is now closer to one in five, a remarkably low number by the standards of most comparable countries.

This may say something about the negative images of parties in the public mind here, since more like 40% will finish up voting the same way in 2007, 2011 and 2016 according to my analysis of RED C polling evidence, but it also suggests that the standing decision that many have is based less on identity, as it once was, and more on calculations of self-interest.

Much of the movement that may take place between now and polling day may also have little impact on the result. In the last few elections, the polls have not indicated any substantial change in party support during the campaign, even in 2011.

In 2016, as in the several elections previously, the election will not happen unexpectedly (as it did in 1989 and 1992 for instance) and so voters have had a long time to think about what they will do.

Even if in the elections up until now one in three voters has "finally made up their mind" in the last few weeks, the net impact of this is no greater than it might be if only one in ten voters had changed their previous inclination – a 2-3% change in support for each party.

Of course, such a small degree of movement could prove to be critical when it comes to the politics of government formation after the election. Add 2-3% to each government party and the government could get very close to the majority it seeks; subtract 2-3% from each and we will be in uncharted territory.

We do have a number of small new parties running this time, hoping to get themselves into the set of options from which voters make their choices, and this could mean we see more change as these parties use the campaign to get themselves heard. Even so, many of their best bets for seats are already well-known incumbents. 

All sorts of issues could prove decisive in generating small net movements in support, and I will be exploring some of these in later posts, but the Government's best hope is that people opt – as they did in the face of impending economic decline in 2007– for a safer set of hands on the economic tiller.

Bertie Ahern and FF were simply trusted more on the economy than the opposition, however weak his government’s record was in all sorts of respects, and the same was true in 2002.

All the available evidence is that the current government wins if economic management is the issue.

The 1997 election seems to have been different in that the FG coalition were seen as better economic managers, but voters opted for FF, whose tax-cut package was more understandable.

By that time anyway, Dick Spring's Labour goose was well over-cooked, despite the rapidly growing economy.

Voters most worried or angry about other big issues  – from health to housing and certainly Irish Water  – and those most concerned with getting a "fairer" society may be captured by the opposition already, but Labour in particular will be hoping to convince a few waverers that it is the best bet available.

None of this is meant to suggest that if one party stopped campaigning now, or its candidates decided not to spend every hour of daylight (and some of the night hours) knocking on doors that it would make no difference, but rather, given that the campaign will take place, the outcome will be relatively close to what the polls are now suggesting.

The key term here is "relatively close". Success in winning the battle for the small margins will make a very big difference to what happens next. 

By Michael Marsh, Emeritus Professor, Trinity College Dublin