It is now rarely recalled how few predicted the hung parliament of 2010. Until just months or even weeks before the election, most pundits and commentators were forecasting a Conservative majority. A survey in April 2010 by the Independent on Sunday of eight polling company heads found that seven expected a Tory majority of between 10 and 50 seats, with just one (Ben Page of Ipsos-MORI) correctly predicting that they would fall short.
Having failed to see the last hung parliament coming, the Westminster commentariat is determined not to repeat this error. This explains why talk of coalitions, and the stance Labour and the Tories should adopt towards the Lib Dems, has dominated conversation in the last fortnight. But after wrongly predicting a majority government in 2010, the press may now be making the reverse error: forecasting another hung parliament, rather than an overall Labour victory.
With 15 months to go, Miliband's party continues to lead the Tories in the polls (as it has done for the last three years) by an average of five points. While the return of sustained economic growth has increased the Conservatives' lead on the economy and improved consumer confidence, it has not changed voting intentions in the way that many expected.
It is true that Labour's lead during this parliament has not been as large as those enjoyed by oppositions in the past (most notably Neil Kinnock's Labour). As former Downing Street strategist Andrew Cooper is fond of pointing out, no party in modern times has gone on to form a government without at least once achieving a vote share of 50 per cent (Labour's highest rating to date is 46%, achieved in a MORI poll in November 2012). But this iron rule ignores the fact that much past polling overestimated support for Labour by failing to account for the "shy Tory" factor (hence the 1992 polling disaster) and that in a four-party system, with UKIP consistently polling around 12 per cent, it is no longer possible to achieve leads of 20 points. But in the case of Labour at least, it remains entirely possible to achieve parliamentary majorities.
As too few remember, when Tony Blair won a third term in 2005 he did so with just 35 per cent of the vote, the lowest share of any winning party in British electoral history. With the boundaries unchanged, Labour could, as one senior strategist told me last year, conceivably win a majority with as little as 34 per cent. In 2005, the party won a majority of 66 seats with a lead of three points but in 2010 the Tories fell 20 short with a lead of seven. This apparent bias has less to do with the unreformed constituency boundaries than it does with the fact that Labour's vote is far better distributed than the Tories' and that it benefits disproportionately from tactical voting.
Uniform swing calculations can, of course, be an unreliable guide to election outcomes since they don't take into account factors such as the incumbency bonus and above-average swings in marginal seats. Had there been a uniform swing in 2010, the Conservatives would have won 14 fewer seats, Labour eight more and the Lib Dems five more. But even if, as seems likely, the Tories perform disproportionately well in their existing seats, Miliband has a significant chance of retaining the lead he needs for a majority. Crucially for Labour, polling by Lord Ashcroft suggests that it is winning an above-average swing in its target seats.
The Tories' fortunes are likely to improve as the economic recovery accelerates and as Labour comes under ever greater scrutiny (hence why I put the chance of a majority no higher than 60 per cent). But even if the opposition's lead were to halve, it could still reasonably hope to emerge as the overall winner. One of the key points in Labour's favour is the unusually low level of switching between the two main parties (just 5 per cent of 2010 Conservative voters currently back Labour), with most of the increase in its vote share due to Lib Dem defectors. Unlike in the past, this means that falling support for Labour doesn't automatically translate into rising support for the Tories. In large parts of the country, the Conservatives simply remain too toxic for voters to lend them the support they need to stop Labour (no matter how strong the economic recovery). As polling published by Ipsos MORI this week shows, 40 per cent would never consider voting for them, compared to 33 per cent for Labour. Miliband is fishing in a far larger pool than Cameron.
At present, the media debate does little to reflect these underlying realities. But as the experience of 2010 shows, the press rarely offers a reliable guide to the result. If 1992 was a pollsters' disaster, then 2010 was a commentators' disaster. 2015 could yet be the same.