Real cost of motoring in the UK 1987-2008

At the moment, there is a lot of howling about fuel prices and how expensive driving is and how the economy will implode and millions will be forced to gnaw their own arms off so they can feed their cars.

In fact, in real terms, motoring has become cheaper.

Since its peak in July 2000 motoring is 14.5% cheaper (compared to RPI) or 13.5% cheaper (compared to RPI excluding mortgage interest). Compared to average earnings motoring is 22.4% cheaper. Over the two or so decades since 1987 motoring costs have fallen by 7.7% (RPI) or 0.3% (RPI excluding mortgages).

graph

Discussion

Source: Office of National Statistics, see below. Units are percentage points, related to the real cost in 1987; for earnings they are arbitrary units normalised to approximately correspond with the other two series in July 2000. I picked 1987 as a startpoint because that's when the ONS's motoring costs data seems to start.

The ONS figures take account of all of the costs of owning and running a motor vehicle, including purchasing, maintenance, petrol and oil, and tax and insurance. See http://www.statistics.gov.uk/pdfdir/cpi0408.pdf, particularly PDF page 20.

See also this BBC news article which has a few inflation-adjusted petrol price charts and some figures (which may have been just plucked out of the air of course).

Methodology

I visted the ONS website and found the following selector for time series data: http://www.statistics.gov.uk/StatBase/TSDtimezone.asp. Then I navigated as follows:

consumer price indices, view tables, go.
2.1: RPI indices 1987 to 2008: Jan 1987=100, view series, go.
CHBK: RPI: motoring expenditure (jan 1987=100), add to selection, go.
CBZW: RPI: all items exc mortgage int payments and indirect taxes, add to selection, go.
CHAW: RPI: all items retail prices index, add to selection, go.
Download as CSV. (Data downloaded 2008-04-30.)

I imported the resulting CSV into OpenOffice, and divided CHBK by CBZW and CHAW. I saved the result and also exported it as CSV. I processed the CSV with a Perl script to remove the annual and quarterly data and the footnotes, and reformat it for gnuplot. I then processed the result with gnuplot to produce the graph you see above.

I separately fetched the earnings data. Starting from http://www.statistics.gov.uk/StatBase/Product.asp?vlnk=550:

Time Series Data - Labour Market Statistics
15: AEI: main industrial sectors including bonuses: SA, view series, go.
LNMQ: AEI: Whole economy SA inc bonus: Index 2000=100: GB, add to selection, go.
Download as CSV. (Data downloaded 2008-04-30.)

I cut-and-pasted the cells into my original spreadsheet. I had to renormalise the data. I chose the July 2000 peak as the renormalisation point; I took the arithmetic mean of the two real-terms motoring costs already calculated and obtained a normalisation factor of 0.6. This makes the figure for my related-to-earnings value roughly level with the other two figures on that date.

The files I created are available for download.

Ian Jackson 30.4.2008; last updated 2.5.2008.