The problem with history is that it cannot make up its mind. Should the Prime Minister call a snap early election? Should he wait until the summer? Should he go the distance and call the election only when he has to, in May 2010? History stands up, clears its throat, then says “yes, no, whatever”, before sitting down again.

The conventional wisdom is that James Callaghan lost in 1979 because he waited too long, while Harold Wilson lost in 1970 because he did not wait long enough. John Major held off until the last minute twice, with widely different results. Clement Attlee dissolved his Government after less than two years and lost; Wilson did the same thing after roughly the same period and won. Gordon Brown is a keen student of the experiences of his predecessors and these will not help him in the slightest.

There will be copious debate during 2009 about the political advisability of an election. And this will not help Mr Brown either. Having been, at one time, 20 points behind the Conservatives, Labour was, in the latest Times/Populus poll, only 4 points behind. Should he consider this a triumph (he has cut the lead dramatically) or a disaster (even now, at the height of favourable coverage for him, he trails David Cameron)?

It is also hard to predict which direction the polls will take next. Is this as good as it gets for Mr Brown? Or does he now have the momentum to take the lead? And if he does, will it last? Unfortunately for him, the moment that he knows the answers to these questions may also be the moment when it is too late to act upon them. And so he is alone with his decision - an immediate election is unlikely to yield a Labour majority, a very late election might be a calamity for him, and the outcome of a constest between those dates is unknown and unknowable.

Perhaps, since it is impossible for him to calculate his own advantage, the Prime Minister might attempt a novel approach. He might try asking a question that previous occupants of 10 Downing Street have shown scant interest in - when would it be best for the British people for an election to be held?

There was a case for Mr Brown seeking a fresh mandate immediately upon taking office. He was right to feel, however, that he was under no obligation to do so. The most important argument then for delaying any contest was that a later election would give voters a chance to judge Mr Brown based on his performance in office. This argument is stronger now.

Some of the issues in the next election can be judged now. The public know enough to be able to give their verdict on, among other things, the handling of the housing price bubble, the reform or otherwise of public services and the record on managing the public finances. Voters will already have a view on the origins of the downturn and the Government's culpability. On the Government's handling of the credit crunch once it began, however, many will want longer to make a considered judgment.

Whether it was sensible to borrow vast sums to cut VAT, whether the bank recapitalisation scheme was well designed, whether aid to industry and mortgage holders was ill advised, all this will be much clearer in a year. Who knows when it will be in Mr Brown's interest to hold an election? But for the voter? 2010.