Time to measure up

Economics:Inflation is much in the news at the moment and has been for some time. There is a strongly embedded perception that it has become a serious and persistent problem in Ireland and that the current inflation rate is running well ahead of forecast and of rates in our main trading partners.

At the same time, there is talk of revisiting the terms of the latest national wage agreement in order to address the supposedly much higher than expected rise in the cost of living since that agreement was reached in early 2006.

On the face of it, there seems to be a solid basis for all this hand-wringing. According to the Central Statistics Office's consumer price index (CPI), the inflation rate in the 12 months to May was 5 per cent, a value that it has stuck stubbornly close to since December.

This represents a marked acceleration from the 3-3.5 per cent rate of early 2006 and is more than double the rates reported in our main trading partners. It is also considerably higher than the inflation rates that were being forecast for 2007 a year ago.

However, this is an area where everything is not quite as it appears. Other measures of inflation paint a different picture. The EU-harmonised index of consumer prices (HICP) - so called because it is based on a common methodology across the EU - puts the Irish inflation rate at just 2.7 per cent in May, not a great deal higher than the EU average of 2.2 per cent.

Moreover, that measure indicates that inflation in Ireland has been broadly stable since early 2006, having fluctuated mostly in the range 2.5-3 per cent. Using this measure, there is no evidence of a recent accelerating trend.

Why are these two inflation barometers producing such different readings and which one is superior? These are questions that it might be tempting to dismiss as obscure and technical. However, the answers are central to intelligent public discourse and sound policymaking.

The measure used to estimate inflation is pivotal to a wide range of economic decisions, including wage-setting and the determination of pensions, social welfare payments and price increases for regulated entities. When different measures yield radically different estimates, choosing between them becomes a correspondingly important exercise.

The items included in the HICP are a subset of those that make up the CPI: the coverage of the HICP is about 90 per cent of the CPI. The most important exclusion from the former is mortgage interest, which accounts for virtually all the difference between the two estimates of inflation.

Hence the annual rate indicated by the CPI excluding mortgages (CPIX) in May was 2.6 per cent, almost exactly the same as the HICP figure.

The corollary, of course, is that the mortgage interest component of the CPI has been rising extremely rapidly - by 47 per cent in the 12 months to May. Simply put, there are two reasons for this rapid escalation: rising interest rates and rising house prices*.

I suspect most people believe instinctively that mortgage costs should be included in a measure of consumer price inflation. However, for a wide range of reasons, economists believe it is wrong to do so.

Perhaps the most telling reason has to do with the fact that a house is an asset as well as the source of a stream of accommodation services. The costs of financing house purchase through a mortgage are usually outweighed by the appreciation in the house's value and, on this account, should not be reflected in the CPI.

On the other hand, the cost of the accommodation services provided by a house should be, and this is done most cleanly by reference to prevailing levels of rent in the rental market.

No other OECD country except Britain compiles a measure of inflation that treats owner-occupied housing in the same way as the Irish CPI. Even in the British case, the measure in question (RPI) has been supplanted for policy purposes by a variant that explicitly excludes mortgage interest (RPIX).

Among OECD countries, the most common treatment of owner-occupied housing for CPI purposes is one based on prevailing levels of rent. The next most common approach is to exclude owner-occupied housing entirely.

Applied to Ireland, either approaches would produce an estimate of the current CPI inflation rate of not much more than half that produced by the present methodology.

Given the huge upward bias in the existing measure and the threat that its continued usage poses to our already depleted competitiveness through compensatory pay claims and inflated growth rates of public spending, there is an urgent need to win acceptance of a sound alternative.

Happily, there are two such readily available: the HICP and CPIX. Of the two, the HICP is probably the better on grounds of direct comparability with the other EU member states.

*See Colm McCarthy's recent paper Owner-Occupied Housing Costs and Bias in the Irish CPI (UCD Centre for Economic Research, WP07/07) for an extended discussion.

Jim O'Leary lectures in NUI Maynooth. He can be contacted at jim.oleary@nuim.ie