In an address to the Patrick McGill Summer School in Donegal in July 2004, Garret FitzGerald, emphasising that he was no longer actively engaged with the Fine Gael party, revealed that he had been surprised at how well it had done in that June's local elections, writes Noel Whelan.
He attributed his surprise, in part, to the fact that national media no longer seemed to cover politics outside the media and government beltway of Dublin 2 and 4, to the extent he felt it once had. As a consumer of political coverage at national level, albeit a relatively astute one, he felt he had as a result been left blind-sided about the extent of the party's improvement around the country since its disastrous 2002 election.
It's a point worth bearing in mind as we head into the last, intensive, phase of this election contest. There are real limitations on the capacity to which any of us involved in covering or assessing the forthcoming contest at national level can identify the true significance of happenings in individual constituencies. In 2002, much of the national media failed to recognise the significance of the relatively rapid rise of some of the new Independent candidates, whose support surged on the back of concerns about hospitals or other local issues.
I've been in different parts of the country over the last week or so and have picked up copies of a variety of local papers. I was struck by three common features in their coverage of the election campaign.
Firstly, the number of local papers with Bertie Ahern on their front page is incredible. His whistle-stop tour is having a significant impact, at least in terms of local media coverage. Enda Kenny has also been travelling around but this does not, as of yet, seem to be reflected to the same extent in local newspaper coverage. A day spent by Bertie Ahern sweeping through two or three counties leaves a significant photo trail: Bertie on walkabouts with local candidates, Bertie hugging babies or their mothers, Bertie listening to a local delegation.
Secondly, the level of advertising in which candidates for this election are already engaged in is phenomenal. The most expensive and effective advertising slots in a local newspaper are the bottom corners of the front page. In almost all of the local papers I came across last week, one or both of those front page slots had been taken by an election candidate. They were in the main personalised adverts, featuring and presumably paid for by one candidate only - full colour, well-designed and focusing on some unique selling point for the candidate. If the local newspapers are anything to go by, then the amount of money being spent by candidates in the lead into this election will be multiples of that spent in any previous contest.
Thirdly, much of the reportage of the election campaign in local newspapers is focused on constituency polls. This takes the form of extensive coverage of polls conducted for the local newspapers, or coverage of rumours or suggestions about what private polls conducted for the political parties are suggesting.
This localised polling has some value, especially if done by one of the reputable five companies. It often reveals an incumbent's vulnerability which might otherwise not have been apparent or can indicate which newcomer is likely to be a real contender. However, constituency polls must also come with strong health warnings.
Most constituency polls - even those conducted this close to the election - overstate the big name candidates in the constituency. Many voters, including those who know which party they are going to support, may not have given much thought to the precise candidate for whom they will vote. As a result, it is often the more recognisable name in the constituency which they choose when pressed by a pollster or forced to complete a mock ballot paper. When the election campaign actually begins and voters become more aware of the full line-up of candidates, some of the support registered for the big names is likely to fall away to running mates or even to other parties. There is an extent to which the published constituency polls are giving some in Fianna Fáil in particular - since it has more high profile incumbents - a false sense of security.
In addition, the usual riders about margins of error must be amplified for constituency polling where the sample is smaller. The extent to which constituency polls can accurately measure the destination of transfers is also severely restricted. Polling agencies are usually careful to warn that later seats cannot be predicted on foot of these polls, but newspapers are often more reckless in suggesting that the poll reveals a certain outcome. When it comes to eliminations or surplus distribution in the mock count conducted by the polling agencies, the number of sample ballots involved is very small and the margin for error on the pattern of transfers is very large.
The better constituency polls can, at most, be a useful indicator of what first preferences - within a 3 or 4 per cent range - a party would get if an election was held on the day the poll was taken.
Indeed, the publication of a constituency poll has an impact on the conduct of the campaign and on voter behaviour in the polling booth. In a study of the 2002 election, several candidates identified the publication of a local newspaper constituency poll as the most significant event in the campaign.
Some who lost out felt party supporters, believing the incumbent to be safe, moved to a weaker candidate whom they were convinced needed to be shored up. Others who won, credit a good showing in a constituency poll as giving their campaign key momentum early on.
In 2007, the shifts in the last weeks of the campaign are likely to be even more dramatic than in 2002 and the need to reflect them in national political coverage will be even greater.