The polls indicate it is reasonable to expect turnout to be up on Saturday -and that may make all the difference, writes Richard Sinnott
At this stage we can be pretty certain about the current overall state of mind of the electorate on the question of ratification of the Nice Treaty. According to today's Irish Times/MRBI poll, 42 per cent are inclined to say Yes, 29 per cent are inclined to say No and another 29 per cent either don't know or will not vote.
These figures are broadly in line with the other polls that have been published and there is not much evidence of a shift in voting intentions since the beginning of this campaign. On these figures, the Yes vote is clearly ahead.
What is much less certain is whether these numbers will translate into a Yes outcome when the votes are counted next Sunday. This is because, on the surface at least, not much has changed.
Today's poll looks very much like the final Irish Times/MRBI poll in the last referendum, when the figures were 45 Yes, 28 No and 27 "will not vote/don't know". And, of course, given those figures, the No side won.
Another similarity with 2001 lies in the evidence in today's poll that the broad measure of attitude to European integration ("unite fully" versus "protect independence"), has, from the pro-treaty point of view, slipped slightly over the course of the campaign - from 46/37 to 43/42.
The No campaign may indeed take some encouragement from the fact that the Irish Times/MRBI polls showed a similar shift in the last campaign. They may take even more encouragement from the fact that back in 1998 and also in 1996 the gap between these two preferences was much greater (in favour of the full integration option - see chart).
But perhaps this kind of shift is just what one would expect in a referendum campaign. The debates raise all sort of issues without necessarily resolving them and the voters respond by running for the cover provided by the "protect independence" option.
It may also be that what is measured by this question is not what a referendum like this is really about. It is quite possible to have a preference for protecting Ireland's independence and still vote in favour of ratification of the Nice Treaty. Much of the Yes campaign has been designed to reassure voters on precisely this point.
What, then, are the issues that have been on voters' minds and do they throw any light on the probable outcome? As the polling companies take different and generally complementary approaches to measuring the concerns of voters, it is useful to combine the evidence from the various polls (Irish Times/MRBI, Star/Lansdowne and Sunday Independent/IMS).
Together these show that the issues of sovereignty (power and identity), economic insecurity, neutrality and immigration are the main concerns of No voters.
There is some evidence that the neutrality issue has increased in salience for No voters since the first Nice vote and even clearer evidence that the issue of immigration has moved up in their priorities.
There is also clear evidence that lack-of-knowledge as a reason for voting No has receded in importance. Finally, all the polls tend to agree on the fact that the growth in disgruntlement with the Government plays only a minor role in the No vote.
On the Yes side, the salient issues are enlargement (the argument that "it's only fair"), expected economic benefits and the desire not to be politically or economically marginalised.
Comparing this referendum with the first one, it is striking that intending Yes voters articulate or accept a wider range of arguments for voting Yes than Yes voters did in 2001.
This suggests that the Yes campaign may have been more successful in getting its message across (or even in having a message) this time around; it may also suggest that, when it comes to polling day, a more firmly buttressed Yes vote may be more likely to turn out.
A probable increase in turnout is also indicated by two other pieces of evidence. The first is that people's confidence that they understand the issues involved in the Nice Treaty has increased appreciably.
Looking at this over the four Irish Times/MRBI Nice polls, the proportion who feel that they have "a good understanding of what the treaty is all about" has more than doubled (from 11 per cent at the beginning of the first Nice campaign to 24 per cent in today's poll). The proportion who either feel they have a good understanding or at least "understand some of the issues but not all" has gone from 37 per cent at the beginning of the Nice I campaign to 47 per cent at the end of that campaign, to 53 per cent at the beginning of the second Nice campaign and now stands at 64 per cent.
Since there is convincing evidence from the last referendum that it was precisely a lack of confidence in understanding the issues that was the major factor leading to low turnout, it is reasonable to expect turnout to be up on Saturday.
The second piece of evidence relates to intention to vote. It must be emphasised that intention to vote is a notoriously bad predictor of the absolute level of voting. However, the evidence can be used to assess in a very rough way whether turnout is likely to go up or down or stay the same.
Putting the evidence of three Irish Times/MRBI polls spanning this and the first Nice referendum, together with the evidence from the Sunday Independent/IMS poll, shows a significant fall (7 or 8 percentage points) in the proportion of people indicating an intention of not voting. We can reasonably expect that voting will increase, if not by this precise amount.
Given the figures in today's poll (42 per cent Yes, 29 per cent No and 29 don't know), what does a possible increase in turnout mean for the outcome? Perhaps the best way to think of it is to regard the poll figures as indicators of latent support for the Yes and the No sides, with an additional reservoir of don't knows to be won over. For each side, however, activating their latent support is as important as, or perhaps more important than, winning over the don't knows.
Last time out and on the basis of a distribution of latent support very similar to that in today's poll, 16 per cent voted Yes, 19 per cent voted No and 65 per cent abstained. In other words, in June 2001 the Yes side was particularly bad at activating its latent support. It might be doing a bit better now.
However, there is no certainty about the indications that turnout may be up or about which side might benefit. Accordingly, the message for the campaigners on both sides is as clear as it is traditional: get that vote out! The message for individual citizens who may be tempted to abstain is also clear: whether you are a No leaner or a Yes leaner, if you do not vote you are letting others (and probably a minority of the electorate) decide an important issue for you.
Prof Richard Sinnott is director of the public opinion and political behaviour research programme at the Institute for the Study of Social Change in UCD.