ANALYSIS:A monthly publication means this will be a more useful measure of house price movements
PROPERTY PRICES. They dazzled this country for a decade, feeding a frenzy and – among many other things – vulgarising dinner party conversation.
Since the crash they have depressed a people and an economy. Until they stop falling the prospects of recovery in the wider economy will be lessened.
On top of this uncertainty has been uncertainty about the reliability of basic property price data. Yesterday’s publication of a new house price series from the CSO will help to bring more certainty to discussion of what is happening to prices.
The new index is the work of professional statisticians. Unlike all of the other existing indices, the sample of properties used by the CSO is not self-selecting, as is the case for other indices produced by property websites and estate agents.
The figures used by the CSO come from eight big mortgage lending institutions. Over the five years for which figures are available (2005-09) house and apartment purchases funded by their mortgages ranged from 70 per cent of the total to 94 per cent (almost all of the remainder were cash-based transactions).
This means the sample size used in the new price index covers a large majority of the transactions, making it far bigger than any that previously existed. This improves the accuracy of the figures in reflecting what is really going on in the property market.
All that said, and in fairness to the other indices, the new CSO figures differ little from the others.
Another important difference with other indices is that the new CSO model is based on the price paid, not the asking price or an estate agent’s valuation.
This is obviously a more accurate reflection of the true picture.
A clear improvement of the new figures on anything previously available is the frequency and timeliness of the figures. All other indices are quarterly. The new CSO figures are monthly and will be
published with a delay of about six weeks.
It is important to state that perfect accuracy in most statistics is impossible to achieve. It is particularly hard in property.
Measuring house price changes is more difficult than measuring consumer price changes.
A bag of frozen peas is identical to another bag of frozen peas. Pea price inflation is measured by taking the price of a bag every month and tracking the change over time.
Houses and flats are infinitely more variable than peas – with size, location and build-quality determining prices. They are also traded much less frequently than bags of peas. As a result, property price indices will always be less accurate than consumer price indices.
When asked by The Irish Timeswhether the CSO thought about seasonally adjusting the figures, the statisticians said they had found no time-of-year effect and, therefore, didn't adjust.
Why has the CSO produced this index now? The answer is Europe. EU member countries have been harmonising the way they compile their statistics for years.
The bloc’s statistical wing – Eurostat – agrees with the states on what figures they should compile. More importantly, they set out how they should be compiled.
This makes them more readily comparable and it lessens the temptation of governments to lean on number crunchers to massage the figures.