Before I remind you of the obvious, let me first boost my credibility for predicting these things (whether by foresight or just plain luck). Two years ago I wrote in this space about most Conservatives at the conference being confident of winning a majority “because a) the Lib Dems are in calamitous retreat, b) Miliband is hopeless and yet will not be ousted, c) the proposed boundary changes will reduce the electoral bias to Labour”

However, I argued, ‘mean reversion’ would prevail: “a) the situation of the Lib Dems is more likely to recover than to worsen; b) Milliband is likely to become more credible not less credible… c) a large number of MPs and Lords will pursue self-interest and therefore the hoped-for redrawing of seats may be tied up in many knots; and one could add a fourth potential for mean reversion: the return of splits within the ranks of Tories on issues such as Europe and the true nature of Conservatism.”

For 2011, at the height of Tory confidence and Labour misery, I think that was pretty good going. You might say the Lib Dems have not recovered, but recall how they held on to their seat in the tricky Eastleigh by-election. Anyway, I argued then that only one of those things needed to happen to deny the Conservatives a real chance of a majority.

Labour are also very unlikely to achieve an overall win. Their present lead is not big enough at this stage of the cycle, and it is notable that in YouGov’s ‘forced vote’ question (‘if you had to choose, which would you prefer: a Cameron-led Conservative government, or a Miliband-led Labour government?’) the answer is a dead-heat. But that is not the most damning fact: worse still for the opposition is how they are rated by the public on the economy: here’s the two-year change:

Asked in April 2010 if the economy would be stronger if Labour won the next election, 36 per cent said it would be, and 40 per cent said it would not be (a net of -4 per cent). A few weeks ago the judgment was considerably worse: 24 per cent said it would be stronger, to 45 per cent who said not (a net of -21 per cent).

Meanwhile the Conservatives improved, from -9 per cent to +3 per cent.

This is a dreadful situation for Labour. The economy is not the only driver of election success, but can we really imagine a big swing to the party which the electorate predicts – by nearly 2-to-1 – will worsen the economy, over a party that it predicts will improve it? Miliband’s new narrative, that only the rich are benefitting from this recovery, is a powerful one, and the initial Tory response was weak; but it is surely not enough.

Even if Labour does manage a 6 per cent lead in 2015, that only delivers a majority if it’s a uniform swing. Swing is never uniform, and an analysis of previous elections shows that there is a significant incumbency effect in seats that change hands when the government changes. My colleague Peter Kellner has pointed out that when Blair suffered a -2.4 per cent shift in 2001, among the new Labour MPs seeking re-election there was only a nugatory -0.2 per cent swing, and Labour lost hardly any seats. According to Kellner, a 6-point lead could still leave Labour 2 seats short of a majority.

I said we already know the outcome of the next election. Of course nothing is certain: UKIP could do even more damage to the Tories than we are already expecting, or Dave could slay the dragon. But again, ‘mean reversion’ remains the most likely scenario: while UKIP will surely continue to rise in the European elections, it’s likely the currently favoured party of protest will fall back at least a little.

Of course I’m not saying everything always reverts, and there is an important, serious long-term trend here: the two biggest parties are declining in their share of the vote. That trend continuing, but in an unspectacular fashion, is what I believe leads us into an extended period of coalition.

So the outcome of the next election is that the Lib Dems will be winners, in power even if they do end up in fourth place.  But it will be a grisly kind of victory, where they are not truly free to choose, under severe pressure to partner with whoever turns out to be the biggest party, whether they like it or not. It could eventually drive them nuts.

So knowing all this, how will it change the character of the campaign? Conventional wisdom is, you should always just play to win outright. Think about any deals later. That surely is the sensible thing – if it were possible. But I don’t think it is possible. There has never been a British election in which the widely expected outcome is coalition. The issue of coalition will be hugely disruptive to the normal process of campaigning, undermining every policy discussion, weakening every message, sapping the energy of the participants and irritating the voters. That in itself could upset the calculations. So I should retreat from the firmness of my prediction: maybe one party or both will decide to shoot the moon.