Mr NAIRN —My question is addressed to the Treasurer. Is the government aware of claims that, according to Treasury household expenditure survey analysis, five times the compensation provided in the government's tax package is required to look after low income earners? What is the government's response to these claims?

Mr COSTELLO (Treasurer) —I thank the member for Eden-Monaro for his question and I acknowledge the great election victory he had on 3 October.

Back in 1992, Labor Treasurer John Dawkins informed this House that the household expenditure survey was not a reliable way to cost all tax changes. It was not a reliable survey to use on that basis. He said:

The treatment of household expenditure survey data is wrong. It is nothing more than a statistical nonsense.

He also tendered in those large documents a Treasury minute to him as the Treasurer saying why the household expenditure survey was not relevant. It is a survey that is done every five years. If you happen during the survey period to buy a lump sum item, for example, it is assumed you buy one every week. If your expenditure in a particular week includes the purchase of a refrigerator, for example, instead of showing expenditure of $100 for the week, you are assumed on the basis of survey data to have spent $800. In relation to self-employed people, it is notoriously unreliable as to their savings ratios. On the basis of the household expenditure survey, some of them have a dissaving at the rate of 300 per cent per annum.

That did not stop the Labor Party in 1998 reversing its position and saying that the only reliable measure was the household expenditure survey and advertising the day before the election saying that Treasury papers would show that the effect was five times greater than the CPI analysis, which was 1.9 per cent. The analysis that John Dawkins said was right in 1992 was the analysis which was used for the 1985 white paper, where Kim Beazley supported the GST.

Today I release all those documents to the House in response to a freedom of information request, precisely the same thing that John Dawkins did in relation to the request by the then member for Flinders. It will show, Mr Speaker, if you look at it on a household expenditure basis, the effect on the cost of living across 11 different household groups at five different income levels. Remember the Labor Party said it would show five times 1.9 per cent, or 9.5 per cent. The cost of living increase varied between 1.3 per cent and 2.5 per cent. Treasury analysis—remember the government said 1.9 per cent and the Labor Party claimed in advertising it was 9.5—was that the all household price change, on the HES based analysis, was 1.8 per cent, lower than the CPI effect of 1.9. It was lower.

Mr Crean —Is it all the documents, Peter?

Mr COSTELLO —We are back to the documents now, are we? No wonder we are back to the documents. I will table that advertisement, Mr Speaker. The reason why the analysis was not put out—and we said it at the time—was that the Treasury had said in 1985 it had advised the Labor Party to do it on a CPI basis; the Treasury had said in 1992, when it modelled Fightback, to do it on a CPI basis and the Treasury had told us to do it on a CPI basis. The Treasury had said it was unreliable. We had it done to make sure the compensation package would exceed the cost on any measure, which is why we chose four per cent, and the all household result was 1.8 per cent. The Labor Party advertising on the eve of the election, I am afraid to say, was totally false and fabricated.

Mr Beazley —Is it tabled?

Mr Crean —Did you table it?

Mr COSTELLO —I table it.