LRM says, "These are the facts. The Labour Party won the 2001 election with a lead in the United Kingdom of 9.0 per cent over the Conservatives, a swing of 1.75 percentage points to the Conservatives. This swing was not reflected in seats because of increased tactical voting by Liberal Democrat supporters in Labour marginals. The Times Guide to the House of Commons lists 25 polls published in the last month of the campaign by what were then the four leading conventional polling organisations. Every one of these 25 polls exaggerated the Labour lead in the actual result. The average Labour lead given by these opinion polls was 17.9 per cent, almost exactly double the actual outcome. ICM was the least inaccurate, with an average lead of 14.5 per cent, Gallop had 16.25, NOP had 18, and MORI 22.4 per cent. They all predicted a Conservative meltdown which did not happen. They may be doing so again. "
I think there are three points worth making here. First the 17.9% lead in the polls was the average over the campaign. The final-week polls, presumably more accurate, put the lead at a more modest 14.5%, with (I think) ICM the best at 9%, Gallup the worst at 17%. This doesn't necessarily change the thrust of his article -- that today's bad poll position can be overturned -- but it makes it much more modest. Second it is seats what matter in British General Elections -- ask the Liberal Democrats. As I have posted earlier, it is very likely that Labour can survive a much smaller lead and still have a large majority. Third, many of those who are only likely to vote will vote, and presumably they are more heavily Labour supporting. Fourth, one must remember some of LRM's earlier forecasts have been a little inaccurate (thought his stock market crash forecast looks better and better by the day). I remember also an article in wrote in The Times in 1996 in which he produced a chart from the OECD showing the miracle of Thatcher in making the UK's economy grow faster than every other in the 1980s -- including Japan! Unfortunately he had misread the chart, which was of the variability of economic growth, not its magnitude. He finally apologised a few weeks later after The Times had failed to print many letters pointing out the error, which any half-intelligent person should have noted from a ten second glance at the data.
posted by Matthew @ 1/02/2003 03:37:00 PM
Comments: 0
