Old-fashioned bank managers, before the age of call centres, used to regard it as their professional duty to give prudent advice to their clients. One of their rules of thumb was that young buyers should not pay more than four times their income for their first purchase of a house. For most of the past 50 years that advice has worked tolerably well. House prices did average about four times disposable income in the late 1950s, 1960s and mid-1990s.

However, there have been four postwar peaks in house prices, in 1948, 1973, 1988 and 2006. At each of these peaks the average price of a house had risen to six times the average disposable income. The first three of these peaks proved to be unsustainable. In real terms, house prices fell back by a third in each of the four-year periods that followed the peaks.

Of course, there were different economic events to help to account for the declines. In 1949 there was devaluation, in 1973 there was the first shock of higher oil prices and after 1988 there were sharply higher interest rates.

House prices are hard to predict; all who try to advise on economic trends have sometimes got them wrong. Nevertheless, the postwar pattern is clear. There is an average price, which is about four times disposable incomes. There are regular peaks that take prices up to six times disposable incomes. So far there has been no occasion on which such a peak has been followed by a further significant rise, and no occasion on which it has been sustained.

The Bank of England, which controls interest rates, is sceptical about the present level of house prices. Mervyn King, the Governor of the Bank of England, says that the bank cannot understand “why house prices relative to conventional earnings are as high as they are”. The bank did not directly target house prices when setting interest rates, but it is bound to take them into account.

The evidence is reviewed by Fred Harrison in the latest issue of Money Week. He quotes David Miles, the chief UK economist of Morgan Stanley, who gives warning that “a sharp fall in real house prices is likely at some point in the relatively near future”. Clive Briault, the manager of retail markets at the Financial Services Authority, gave his warning to the British Bankers Association that the banks should now factor in the possibility of a 40 per cent fall in property prices. That is not a prediction but a precaution.

The United States has had a similar boom in house prices. There, the peak seems already to have been passed. The recent drop in the construction of new houses in America was the biggest fall since the early 1990s.

Most people, including many economists, do not properly understand how expectation affects markets. It is the market traders who follow the markets’ fluctuating moods of hope and fear. These moods do not alter as though the market were controlled by a dimmer switch which could be moved by subtle degrees.

Expectation is like a rusty, old-fashioned on-off switch that requires increasingly strenuous pressure before it will shift. When it does shift, the sun stops shining. One day the market is full of buyers with no sellers; the next day it is full of sellers with no buyers. That is how one gets the black days.

When the housing market does turn down, the first consequence is that people cannot sell their houses. It is not that they cannot get the price they used to think their houses were worth. The problem is that no one will even come to view them. In the stock market there is usually some liquidity. In the residential market there can be a lot of houses for sale, a lot of debt on those houses, a lot of sleepless nights and no buyers. Apart from anything else, house prices have continued to rise into 2006 because of the expectation that they would go on rising in 2007 and 2008. That expectation may still prove correct, but at some point the music will stop.

The cycle of house prices is related to the political cycle. After 1948, Labour went out of office in 1951; after 1973, the Conservatives went out of office in 1974; after 1989, the Conservatives had a reduced majority in 1992 and a landslide defeat in 1997. Voters do not like falling house prices. Some of them even have their houses repossessed — as many people did in the early 1990s.

This will face Gordon Brown with a difficult question when he becomes Prime Minister, presumably next year. Should he follow the example of Anthony Eden, who went straight to the country on succeeding Winston Churchill in 1955? Eden increased the Conservative majority. Should he follow the example of John Major, who went the whole five years both in 1992 and 1997, once to victory and the second time to defeat?

I think that Mr Brown’s best chance would be to go straight to the country in the hope that he would have some lift in the polls as a new Prime Minister. The country is bored with the Labour Government and that boredom will tend to increase if he waits another three years.

A snap election in 2007 might well be Labour’s best chance of retaining an overall majority. It might also be the last time when the housing market, which affects most voters, is still relatively healthy.

I do not expect Gordon Brown to take this advice. At present the polls give a lead to the Conservatives, ranging between 2 to 8 per cent. Before deciding to call an election any Prime Minister would want to have a consistent lead for his own party, preferably a 5 per cent lead for a six-month period. Before Tony Blair retires, the May elections in Scotland may have further weakened Labour confidence.

All his life Mr Brown has wanted to be Prime Minster. If he becomes Prime Minister in July 2007 he can be more or less sure of three years in office — years in which he can impose many of his ideas. He will not give that up for the uncertainty of an election in September 2007. But by 2010 house prices may have fallen by a third and Labour may lose by a landslide.

Times Recommends

• It’s the beginning of the affair
• Afghanistan: still a price worth paying
• Heroes of the downtrodden