While I’m on the topic of tax and public expenditure, this is a good opportunity to refute some spurious claims made by Peter Saunders of the CIS in a piece in the Oz recently. This isn’t too hard, since the claims have been made quite a few times before, and I’ve refuted them quite a few times previously, a task that becomes a bit frustrating after a while.
First, Saunders says
The top tax bracket was worth 15 times average earnings in 1960; nine times average earnings in 1970; three times average earnings in 1980; but is just 1.3 times average earnings today.
As I pointed out in my 1998 book Taxing Times
The tax reforms of the 1980s abolished the old top tax bracket, for which a marginal tax rate of 60 per cent was applied. An inevitable effect was that the former second-highest rate now became the top rate, even though it was lowered from 49 per cent to 47 per cent. Hence, even though taxes on high and middle incomes have been cut, the top marginal rate is now reached at a lower income, expressed as a proportion of average weekly earnings.
Even sillier comparisons are made with the tax system applicable in the 1950s. For example the Treasurer, Mr. Costello argues for reform on the basis that, in the 1950s, only people with incomes equal to 19 times average earnings paid the top marginal rate of taxation, whereas today people with incomes equal to one and half times average earnings pay the top rate. The ratio of tax revenue to GDP has, of course, risen since the 1950s, in Australia as in every other developed country. But the dominant effect arises from reductions in the progressivity of the income tax system. The current income tax system contrasts dramatically with that of the 1950s which had 29 brackets and a top rate of 67 per cent. Obviously in a system with so many brackets, very few people will have in incomes in the highest bracket.
The Parliamentary Library has a useful article on this topic, which shows how the increase in proportions paying the top rate was mainly driven by the reduction in the number of brackets. More detailed info is available from NATSEM (PDF file). Combining this with AWE data, I estimate that the 47 per cent rate applied at around 1.2 times Average Weekly Earnings in 1974 and about 1.4 times average weekly earnings in 1984, scarcely any different from today.
The main difference was in the very high rates applying to high incomes (the real target of Saunders’ concern, I suspect). The top rate in 1974 was 67 per cent, and it did not apply until income reached 40 000 per year, 5 times the threshold for the 48 per cent rate. Upper-income earners were taxed much more heavily in the past than they are today (though of course it’s necessary to take account of avoidance and evasion when considering actual rates of tax).
Second Saunders compares rates between countries saying
Not only does our top rate cut in too low – the rate itself (48.5 per cent including the Medicare levy) is much too high. In Britain it is only 40 per cent; in Germany it is 45 per cent (and coming down).
But this ignores the impact of payroll taxes and social security contributions. Australia has low payroll taxes and no separate contribution to social security[1]. By contrast, in most of the other countries mentioned by Saunders, these taxes are large. In Germany, for example The employer and employee each make a contribution of 21% (2004)This point has been made so often it’s hard to believe that Saunders is unaware of it.
fn1. It’s also necessary to take account of compulsory superannuation contributions in Australia. But unlike social security schemes in other countries, this is pure forced saving – there is no pooling of contributions in the accumulation schemes that are now standard (except for politicians). So although there’s compulsion used here, it should not be regarded as taxation.