emperors_new_clothes_2023_31Jan2024_smaller
ContactBlack2ContactBlack2
SiteMapBlack2SiteMapBlack2
RPI_minus_CPI_since_1975
[uk_rpi] [why_bother_trying] [RPI_minus_CPI_since_1975]

Although the RPI could have been repaired to make it compliant, it seems pretty certain that this is not going to happen. Instead, it seems that it will, effectively, be replaced by the CPIH rather than CPI, both of which are similar sorts of measure,  compiled under an EU formula. In fact, the figures indicate that the similarities between CPI and CPIH seem to be greater by far than the differences.

The CPI excludes Council Tax, mortgage interest and some other housing costs. On the other hand, the CPI includes some items which are not in the RPI scope, such as charges for financial services. The
CPIH is defined as the CPI including owner occupiers' housing costs. Whereas the RPI uses arithmetic means to combine prices, CPI and CPIH mainly uses geometric means for combining prices. Effectively, the CPI and CPIH cover a broader population sample than the RPI. For given price data, the CPI and CPIH inflation rates are almost always lower than the RPI inflation rate.

Two charts are attached, showing the differences between RPI and CPI between 1975 and 2023 (December to December). The first shows the relative increases (RPI minus CPI) for single years and the second shows the differences by duration (average and standard deviation). It was expected in 2012 that the difference would increase to around 1 % pa but there is still a  relatively steady difference of around 0.7% pa.

Comparing RPI with CPI over 1 year, the average has been 0.79%, with a standard deviation of 1.06%, a minimum of -2.74% and a maximum of 2.93%. Over just 1 year, it is not unreasonable to suggest that the expected difference could be as high as 1%.

However, comparing RPI with CPI over 15 years, the average has been 0.73% (very similar to that over 1 year), with a standard deviation of 0.11%. A value of 1% would be 2.59 (2.84 a year ago) standard deviations away from the mean, which still corresponds to a likelihood of around 1 in 200, namely highly unlikely. Hence, 1% really cannot be taken as a best estimate.