Calling a general election is one of the trickiest decisions any prime minister faces, because it can prove fatal. Get it wrong, and the leader who has miscalculated is likely to be out of a job before long.

Prime ministers obviously want to be sure that their party has a good chance of winning a working majority. That is why, despite this week’s speculation, the odds are still against an October election.

Of the 26 elections in the past century, nine came at the end of wars, the collapse of a government or when the parliament had run its full course. Of the 17 discretionary elections, prime ministers made the right decisions in nine by either increasing or maintaining their Commons majorities. Three were draws.

But five turned out to be failures: in 1923, when Stanley Baldwin lost a big overall majority just a year into the life of the parliament (though he won an even bigger margin a year later); in 1951, when a worn-out Clement Attlee called an election that he might have won later; in 1970, when Harold Wilson misinterpreted an upturn in Labour’s poll ratings and lost to Edward Heath; in February 1974, when Heath called, and lost, the “who governs Britain” election; and in autumn 1978, when James Callaghan failed to call an election that he might have won – or at least held on as a minority government – only to lose badly the following May.

The historically-minded Gordon Brown is well aware of these precedents. Not surprisingly, he is making preparations and looking at raising funds, which is very necessary given Labour’s still huge debts. Admittedly, a lack of finances has never deterred a party leader from calling an election that looks winnable; donors have always produced their cheque books when needed. However, to be effective, money is needed well before a campaign starts, rather than during the final battle, to allow parties to recruit staff and strengthen their party organisation. That is why much of the £2.4 million donations from Michael Brown in 2005 came too late to make a real difference to the Liberal Democrats.

But what are the pros and cons of an October election? The temptations are obvious: riding a surge in the polls and winning a mandate in his own right. Mr Brown cannot be certain that the economy and his ratings will be so favourable next year. On the other side, of course, the Brown bounce could be ephemeral. As I discussed yesterday, such post-handover boosts can be very short-lived, even a mere few weeks. Anthony Eden got in just in time in 1955. A small swing back to the Tories and Labour’s overall majority could easily disappear. The Conservatives hope that calling for a referendum on the European treaty would mobilise sceptical voters. But this has never been a vote winner in the past. Moreover, an October election would come before the end of the inter-governmental negotiations, giving Mr Brown a chance to defuse this issue.

The political scene could look very different after the holidays, as it often does. The Brown team will follow every poll and every focus group in September to assess the strength and sustainability of the bounce. Managing expectations will be crucial to prevent either an unstoppable bandwagon or a sense of letdown. Remember, however, that Mr Brown has always been very cautious and risk averse.

Times Recommends

• Graveyard of empires could claim more careers
• Would you live on the wrong side of the Berlin Wall?
• At last we’re getting our heads out of the sand