Even a stopped clock tells the right time twice a day
Over on political betting Mike Smithson has a post on whether fuel prices drive voting intentions, accompanied by a graph showing that a peak in petrol prices in Summer 2008 co-incided with a low point in Labour's polling, while a slump in petrol prices co-incided with a comparatively good polling position for Labour in January 2009. In the past I've had comments here from other people seemingly convinced that petrol prices are the sole driver of voting intentions.
Below is the same graph of average petrol prices, but overlaid with a line showing the Labour government's lead in the polls (or in most cases, their deficit in the polls) throughtout the time period of 2007 to 2010, rather than just those two points. You can see the peak and trough in summer 2008 and Jan 2009, but you can also see the lack of a relationship the rest of the time - petrol prices dropped sharply in autumn 2006 with no corresponding increase in government support, rose throughout 2008 with no obvious drop in government support, from spring 2009 to spring 2010 there was a steady increase in petrol prices and government support.
There will be some indirect relationship between fuel prices and government support, simply because petrol prices are strongly related to the price of oil and the oil price is plugged into the wider global economy. The collapse in petrol prices between Summer 2008 and January 2009 came as a result of the economic crisis and a sharp retraction in demand for oil. The increase in support for the government probably came from their response to the crisis and a positive public reaction to the handling of the bank bailout - so the correlation between falling petrol prices and rising government support in 2008 probably wasn't co-incidence, but it's likely that the two events shared a common cause (the credit crunch) rather than any direct causal link. Outside that time period, there is no obvious correlation.
