Speculative drilling

For any lecturers looking for a question to go on their undergraduate econometrics exam, Crikey provides some excellent raw material.

15. Do petrol prices fuel political popularity?

Richard Farmer writes:

In the United States, where some of the pollsters track the standing of politicians on a daily basis, the pundits reckon there is a correlation between petrol prices and the standing of the President.

The popularity of George W Bush began its slide to an historic low for a president not because of some particular dislike of the war in Iraq but because of petrol prices reaching a record high partly as a consequence. Now that world oil prices are on the way down, the presidential rating is rising inversely.

In the Australian data I have looked at, the correlation between petrol price and political support is not as clear as it apparently is in the US but it does seem to exist to some extent. (A correlation of -.49 for those who are statistically inclined.) In the graph below, support for the Coalition as measured by Newspoll is on the left-hand index and the unleaded petrol price in cents on the right.

It will be interesting to see what happens over the next month or so. Newspoll this week had Labor again on a two party preferred vote of 53% to the Coalition’s 47% but the petrol price has recently dropped quite dramatically to as low as 110 cents a litre in Sydney.

If the United States experience holds here, an improvement in support for the Coalition can be expected in the absence of some other startling development.

Incidentally, it turns out that my blog hits, which have been steadily rising over this period too, provide an even better fit (correlation=0.59).

This entry was posted in Economics Generally. Bookmark the permalink.

3 Responses to Speculative drilling

  1. Sinclair Davidson says:

    So it’s all your fault. People driving to internet cafes to log onto your site, increasing the demand for petrol and oil. Keep up the good work!

  2. A nice reformulation of the old point about correlation not being the same as causation.

  3. Andrew Leigh says:

    Andrew, like most people, I only have one idea – I just work on finding new ways to express it.

Comments are closed.