Does uprating by prices lead to better results than uprating by wages?
In my earlier post on public sector pensions I asserted that indexing by earnings produces better results than indexing by prices. In the comments Luke Snell queried whether this was still true. He asked whether “this assertion is valid given trends over the past 10-15 years?”.
Good question I thought. So I’ve knocked up a very quick spreadsheet that worked out whether you would be better off with a pension (or anything else for that matter) uprated by earnings, RPI inflation or CPI inflation.
This is it.
|starting year||earnings compared to RPI||earnings compared to CPI|
It shows that something uprated by earnings from 1970 would now be 79.8 per cent higher if uoprated by earnings rather than RPI inflation. (The CPI measure only goes back to 1988, so only more recent figures allow a CPI inflation).
What the figures show is that you do better with indexation to earnings starting in all the years shown other than 2005.
It is of course possible that this recent trend will continue into the future, but I think few would agree with that – especially since the government has shifted to using the generally lower CPI inflation measure, which the OBR think will be 1.4 per cent lower than RPI in future.
Of course the shorter the period studied the higher the risk will be that the result will not be typical. Inflation has been atypically high over the last couple of years so while earnings growth has slowed over time, we cannot assume that prices will continue to be high over the decades long period over which pensions accrue.
For those intererested in the technical details, I’ve used the September figure for change over the last twelve months for RPI and CPI. I’ve alse used the September figure for earnings, but used the Average earnings index up to 2000, but the avereage weekly earnings since 2000 as that is the figure the government now uses to measure earnings increases, but as it was not published before 2000 I’ve used the AEI figure.
I’m aware that this is probably not the best way to measure the difference between these indices as it would probably be better to produce an average annual figure for the difference between earnings, CPI and RPI over different periods. Comparing 1970 with 2000 in the chart above is not very revealing as there is 30 years more uprating in the 1970 figure. I suspect someone who knows more stats than me would use a geometric mean for this.
But another way of putting this is that if we compare earnings with CPI inflation, earnings have been higher than CPI for every year bar five since CPI started in 1989. But four of those years were the last four. (The other was 1993.)