It ain’t so

Andrew in the See how well the government is going post points to data revisions suggesting that they ‘contaminate’ the data found in the Budget papers.

Comparing the taxation and spending as a % of gdp of the Howard and Rudd/Gillard Governments is not really a fair comparison. The first budget handed down by the Labor Government was the 2008-09 budget. This was when the accounting system was changed by the ABS and Treasury, different to the accounting during the Howard Government.
…
That source shows that due to accounting changes, the gdp has increased according to the books above what the economy grew to therefore making Labor’s taxation and spending look smaller than what they were in comparison to the Howard Government. Some estimates state that this change of accounting may have increase the gdp measured by the ABS and treasury by 4.5%.

http://barnabyisright.com/resources-articles/labor-fakes-gdp-by-4-5/

It is true that GDP has been restated and comparing data from different budget papers (pre and post the restatement) would be inappropriate. The other change that occurred with the change in government is including the GST in the Receipts time series. The thing is Treasury and/or the ABS backcast the GDP adjustment so all the data has been updated. But Andrew goes beyond that pointing to a figure at the Barnabyisright site. Alx makes that point too. They suggest, following Barnabyisright, that the Howard era data was singled out for ‘special treatment’ of some sort – and the graph seems to support that view.

Also, regarding the back casting of GDP – the revisions conveniently appeared to do a hockey stick in 2007 (previous commentor’s link to barnabyisright website).

As much as I love a ‘pollies and bureaucrats up to no good’ conspiracy, what we’re looking at there is a size effect. What has happened is as follows: somebody took the data from the Appendix of the Budget Papers reversed out the implicit GDP figure pre and post the restatement and subtracted one from the other. Then graphed it. The difference has grown over time … because the economy has grown over time.

So I went to the original source that Andrew linked too. Original data is here and revised data is here. Yes, I know, that should be “are” here. Anyway I then compare the two series, just looking at the Receipts to GDP ratio. In this graph I just show the Howard era and the current government.

It looks more or less like a parallel shift down. I then calculate GDP implied by those numbers, calculate the differences, and also the percentage change. The differences look very similar to before (note I have a longer time period).

But what does the difference look like as a proportion of the original GDP figure?

Fairly flat. Even if you didn’t want to argue that it was flat and ‘grew’ in the last few years of the Howard era, that makes the Howard government look better not worse.

Now I have done this for Payments and Cash Balance and an average of the three variables too. I did that to iron out any rounding errors that may creep in. It is hard to get excited about this – it is slightly annoying if you want to compare forecast across Budget Papers (I did want to do that once and got very frustrated). The only area that looks anything close to problematic is the cash balance – there the numbers don’t change much, but the figures are small and the rounding errors likely to be large, so I’m not worried there either.