The Royal Statistical Society has been taking an active role in the continuing debate surrounding the way in which inflation is measured.
In a recent Financial Times article
, economics editor Chris Giles cast doubt on the Carli index, which is used to calculate the average price of a sub-set of items in the retail price index (RPI). He cited it as the main cause of the increasing disparity between RPI and the consumer price index (CPI). ‘Every year the Carli index remains part of the RPI calculation, it imposes a tax of a little under £1bn on society to give windfall benefits to the holders of index-linked government debt,’ Giles warned.
However, the RSS responded to the article urging caution against ‘an unsatisfactory quick fix’ and arguing that the current major research programme into the formula effect being carried out by the ONS must be allowed to run its full course. ‘CPI also lacks public confidence,’ vice president Jill Leyland pointed out in a letter also published by the FT
. ‘It is not seen to reflect the typical consumer’s experience, partly due to the formula effect issue, but also because it does not include owner-occupier housing costs.’
Tony Cox, chair of the RSS Statistics User Forums, RPI CPI User Group, also responded to the FT article
, saying that the formula effect ‘is just one of a number of important differences between the UK’s two principal indicators of inflation that deserve equal attention’. He also referred to taking account of owner-occupier housing costs in a new consumer prices index as an important change currently being considered by the Office for National Statistics.
The RSS held a meeting last week on the question of how to include owner occupier costs in the CPI. The meeting was arranged at the request of the ONS as part of its major consultation on the issue
. Attendees heard, and enthusiastically debated, the case for and against both the method recommended by the Consumer Prices Advisory Committee and the alternative option. Presentations
can be seen on the RPI CPI User Group section of the StatsUserNet
site where there is also lively discussion of the issue.
The UK Statistics Authority is due to make a decision in September.