Polls are the key and that advances in technique and technology makes them if not infallible, then certainly extremely reliable. Trends can be spotted over a series of opinion polls—and just as importantly make the very odd ‘rogue’ poll that is no more than a statistical quirk easier to pick up. Best suggestion is that some of the polling firms are more reliable than others—in broad terms ICM are probably the most accurate and MORI the firm whose figures I would place least confidence in.


But, however you view the polls, the reaction of the firms is the key to making money. As with any in-running event, both market-makers and punters have a tendency in politics just as they have in sport to over-react to the latest event or piece of information. All the evidence suggests that unless it is a cataclysmic Watergate scandal, then internal party rows, the minutiae of policy detail and the latest media froth have very little effect on how people actually vote.


If 1997 saw a base change in the political scenery, then at the time of writing (early 2003) it is likely to take another monumental turnaround to see a change in government in the immediate future. Another system brought to my attention concentrates on certain American states that history has shown to act as a microcosm of the American electorate. Tim Shipman, deputy political editor of the Sunday Express, has researched the voting tendencies of all 50 states and their accuracy in forecasting the overall election result since the Second World War. Notably, no matter what the state of the economy or fortunes of the United States in war, several states are in tune with the common mood and reliably predict the vote of the country.