This morning it was announced that inflation in the Consumer Prices Index (CPI) rose to 3% in the year to April. The BBC interviewers have been asking whether this means that interest rates will rise. It does not. I was among those who, last autumn and winter argued against cutting rates on the grounds that it would be better to fight one macroeconomic battle at a time, seeing off inflation first (preferably without having CPI inflation exceed the 3% threshold) and then cutting rates aggressively from late this year (and also arguing that interest rates were the wrong instrument to use to try to address the credit market problems). This was not the path chosen. Instead, the Bank preferred a middle way: to cut a bit and let inflation go a bit above 3%. I still think that this was a mistake, but that boat has sailed, and there is no question of rates being raised now unless inflation goes above 4% and seems to be still rising (which, though not impossible, still seems unlikely). What we will face is the problem I was concerned about: that we will be unable to cut rates as the economy slows very abruptly in the second half of this year.
With CPI inflation likely to lie well above its 3% threshold for some months, it is perhaps an opportune time to remember what CPI is. "Inflation" is both a very precise term and a very imprecise one. Precisely, it means the percentage rise in a given index. Imprecisely, it relates to something much vaguer: "how the cost of living is rising". A key point to bear in mind is the distinction between a "policy index" and a "cost of living index". The Retail Price Index (RPI) is a cost of living index - its aspiration is to capture an index of how the goods and services people buy rises over time. The Retail Price Index excluding Mortgage Index Payments (RPIX) was the policy index used until 2003. Everyone understood that RPIX was not intended to capture the actual changes in the cost of living. It was just the index the Bank of England used in managing the economy. No-one would have suggested raising benefits or pensions by the RPIX, or that wage inflation in the public sector should be constrained only to rise according to RPIX.
Now in 2003 we switched from targeting 2.5% inflation on the RPIX policy index to targeting 2.0% inflation on the CPI index. This was done for purely political reasons, as part of the UK's assessment of whether we should join the euro. One of the notorious Five Economic Tests related to convergence, Gordon Brown adjudged that we failed the test, and he proposed that we try to enhance our convergence with the Eurozone members by targeting the same index (CPI - which the ECB calls HICP, but it's exactly the same thing). I wrote several critiques of this change, persuaded Michael Howard's team to oppose this change, and, perhaps emboldened by the political debate, Mervyn King joined the growing chorus of opposition.
Why didn't I like the switch to CPI? There were two key reasons. The first, and most important, was that (unlike RPIX) CPI contained no housing costs element at all. Housing costs at the time comprised some one fifth of the all-items RPI. This meant that CPI was not a cost of living index, but a policy index. Furthermore, the exclusion of housing costs meant that Monetary Policy Committee (MPC) decisions would not take direct consideration of effects upon house prices. Third, I was concerned that, since benefits and pensions etc. would continue to be indexed by RPI (quite properly), and since RPI would continue, quite obviously, to be a superior cost of living index, CPI would not garner public confidence - the concern would grow over time that the MPC was not targeting a measure that was relevant to people's lives.
Gordon Brown appears to me to have made the problems I identified worse by doing something I had not anticipated. It seems to me that he has never really understood or accepted that CPI is a policy index, not a cost of living index. Thus, he continually quotes CPI inflation figures as if they reflected cost of living changes. Perhaps this might be written off as good politics, but he has gone further, appearing to use his CPI target as an implicit basis for the control of public sector pay in the past year or so - constraining a number of public sector worker pay rises to 2%. This is deeply confused and counterproductive. It isn't necessarily a problem to employ a policy index that is not a cost of living index - RPIX was very successful. But you have to understand that that is what you are doing. I'm not convinced that he really does.
I would urge that the next Conservative government employs a UK price index (not an EU one), ensures that that price index includes a housing costs element, and is very clear with the public (and in its own mind) about the distinction between its policy target and the actual change in the cost of living.
