The primary objective of the ECB is to maintain price stability, and it measures price stability in terms of the Harmonised Index of Consumer Prices (HICP). As the main European indicator of inflation, the HICP is therefore an extremely important policy variable.
the proportion of owner occupied dwellings differs dramatically across EU countries: for example… only 40% of Germans live in owner occupied dwellings while about 85% of Spaniards live in owner occupied dwellings. Thus omitting owner occupied housing from the HICP will tend to make the indexes incomparable across EU countries.
Incorporating OOH is presently the top statistical priority for improving the HICP. Three different methodological choices are on the table: the imputed rent approach, the user-cost approach, and the net acquisitions approach.
Eurostat is leaning towards the latter approach as being most consistent with its overall HICP methodology.
An OECD working paper from in 2006 concluded that incorporating OOH (using the net acquisitions approach) into the HICP meant a difference of 0.3% between the two different inflation rates, or 15% of the total. This means that over the time the two measures would be expected to diverge. A related working paper predicted that this effect would intensify if housing prices continued to rise dramatically.
Statistics Norway performed a similar exercise for a period of sharp housing increases and found a difference of 0.5% per year between the HICP rate and an HICP rate incorporating OOH using the net acquisitions approach.
If Eurostat were to incorporate OOH into the HICP using its preferred approach, we would expect a different picture of inflation across Europe over the past decade.
The differences between measures of inflation open up a huge set of issues in terms of wage negotiations at the national level, as well as in terms of measuring competitiveness gaps and coordinating wages across Europe.
Here in Spain, the HICP is useful for comparisons so long as we are prepared to ignore the housing bubble. Otherwise, we would expect Spanish inflation to have been systematically understated for a decade. The Spanish national statistics agency’s housing price index shows an across the board increase in housing prices of 9.8% (11.9% for new housing and 8.2% for second-hand homes) in 2007, the final year of the boom.
A modest upward revision in Spanish inflation would mean that real wages fell in Spain over the last decade. The ILO’s Global Wage Report shows the clear decline in German real wages over the same period. This helps clarify the distributional consequences of the first decade of the Euro: workers across the continent were losing purchasing power.
The global financial crisis has taught us that central banks that ignore the volatile market for housing behave irresponsibly. Now that the bubble has burst, deflationary rather than inflationary pressures from the housing market are not being taken into account.
Continuing to exclude OOH from a policy variable as key as HICP looks to be a bad idea. If Eurostat finally adopts the net acquisitions approach, this will oblige a revision in the numbers as well as a corresponding revision of our interpretations of reality.
 This series only begins in 2007.
 The purchasing power of pensions would also be revised downwards, and etc.
 I’m reminded of Alexander Gerschenkron’s research into how Soviet planners were able to overstate production by tweaking the price indices.