The political economy of unindexed income tax brackets

The survey of opinion amongst Australian Economists made for some interesting reading for me. I found that I where a clear majority of respondents agreed or disagreed with a statement I did as well, and where they were divided, I also had reservations. I guess this means I ‘m safe from the vapid contrarianism and I should start worrying about groupthink instead ¹.

I disagreed with the majority in two circumstances, but neither were on pure economics. One was quotas for female membership on corporate boards. I do think that quotas is a very crappy way of selecting board members, it’s just that I think existing board selection methods are equally as crappy, perhaps worse, so a quota is unlikely to be detrimental. The second was the opinion that income tax thresholds should be indexed to inflation.

The principle of stable and predictable real tax thresholds is sound, but in political practice it can come up against the desire for a cyclically balanced budget – a principle that also has overwhelming support. Forced to choose, I’d go for the latter.

The main reason is that politicians find nominal tax cuts far more attractive electorally than nominal tax increases, and cuts are very difficult. The travails in the US are a great reminder of this. A phenomenally large proportion of the deficit there are ostensibly “temporary” tax cuts from the Bush era, but simply doing nothing, and allowing a return to normal rates is apparently politically impossible. If revenue is ratcheted down, and spending is ratcheted up, the long term stability of the budget is unsafe. Bracket creep at least addresses the former. It allows taxes to be raised automatically, without the opportunity for grandstanding, and to offset the unoptimal choices grandstanding favours.

This came up recently in the comments of Peter Whitefords’s latest post about the tax compensation for the carbon pricing scheme.  The concern raised was that should the elasticity of demand of carbon be higher than expected, the scheme would be too successful. The less than expected revenue from permit sales would not make up for the foregone taxation revenue and the long run balance of the budget would be imperiled. This however would also imply that the price level would not be changed and the tax cuts would no longer be necessary as compensation. Peter was satisfied that bracket creep could be used claw back any proportion of the tax cut that proved unwarranted. I’d add that any compensation that was warranted would come back in the form of ritualised nominal tax cuts – one of the great benefits of incumbency.

There is the danger that the government will find ways to spend whatever money comes in, which is intuitive, and it is something associated with the last years two terms of the Howard Government (baby bonuses, tool boxes, flagpoles etc). It pays to look at this process though. Since tax cuts are politically palatable, many “electoral bribes” will take the form of tax cuts as well as spending. Lets look at constant real income over the noughties. I’ll use the average taxable income of $33093 from 01-02. This income, when kept constant in real terms, remains within the 30 cent bracket, so changes in tax paid are related only to a lowering of the real thresholds due to inflation and an increase in them due to budgetary policy².

There are a number of caveats that are largely irrelevant. Company tax increased elsewhere, as did indirect taxation. Real incomes were not constant. These are interesting in terms of relative tax burden or size of government, but irrelevant in terms of the effects of indexation.

On this crude examination it looks as if the extra revenue from bracket creep was not simply absorbed. The early gains (which possibly was clawing back unwarranted GST compensation) are more than offset by threshold increases later on. It’s tempting to see non-indexation as only a partial sandbagging against Howard’s irresponsibility- that bracket creep couldn’t even keep things up where indexation would have them. Furthermore to have much of the surplus that orthodox Keynesianism required. It’s also possible to make a case that indexation contributes to automatic stabilization; If fiscal contraction is needed most when inflation is rising, bracket creep is most potent just at the right time. I wouldn’t push these too hard though.

I’d certainly prefer adult politicians and media that would allow proper discussion of fiscal balance. In an imperfect world though this might be the best option.

¹ The contrary sentiments in each case are hardly unheard, so I don’t think this is a great risk.

² The Medicare levy is disregarded, and the low income exemption/surcharge thresholds are not relevant at this income.

1. fn1[↩][↩]
2. fn2[↩][↩]