More reform threatened

The Turnbull government is apparently keen to undertake more reform, in this case applying to the welfare system. This weekend I saw a couple of media reports, unsourced but apparently based on material supplied by Social Services Minister Christian Porter, based on the “alarming” projection that welfare payments will rise from $154 billion to $270 billion in a decade, more than can be accounted for by inflation and population growth.

A huge dollar figure makes for a scary headline but not for sensible analysis. The prosaic facts underlying this projection are
(a) Welfare payments are dominated by the old age pension
(b) The value age pension increases in real terms over time, since it is linked to earnings
(c) Even allowing for an increased age of eligibility, and more reliance superannuation, the proportion of the population eligible for the old age pension is unlikely to fall

Put these facts together, and it’s virtually certain that the welfare bill will grow broadly in line with national income (usually measured by GDP), and therefore will outstrip growth in population and inflation. Indeed, a quick check suggests a compound rate of growth of about 6 per cent*, almost exactly in line with nominal GDP (assuming 3.5 per cent real growth and 2.5 per cent inflation).

The government’s proposed reforms appear to involve attacks on recipients of welfare in the popular media sense of the term (Jobstart, disability benefits and so on). But the only ways in which the growth rate of the welfare bill can be held much below that of GDP are
(i) Freeze the real value of age pensions, as has been done to Jobstart since the 1990s; or
(ii) Reduce access to the age pension, either by further raising the access age or by tigntening means and assets tests.

* Porter’s briefing apparently suggested 7 per cent, which I can’t replicate. Feel free to check.

Do we need a global tax to stop rising inequality (crosspost from Crooked Timber seminar on Piketty)

One of the more depressing features of Capital in the 21st Century is the air of inevitability attached to the much-discussed r > g inequality. This is exacerbated, on the whole, by the fact that Piketty’s proposed policy response, a progressive global tax on wealth, seems obviously utopian.

What about a much simpler alternative: increasing the rate of income tax applied to the very rich, and removing preferential treatment of capital income? Piketty’s own work with Saez yields the conclusion that the socially optimal top marginal rate of taxation, after taking account of incentive effects, would be 70 per cent or more. Such rates prevailed, at least nominally, in the mid-20th century, without obvious ill effects. Again, Piketty provides the relevant evidence.

So, is there something about a globalised world economy that renders a return to high marginal rates of taxation impossible?

Read More »

Bracket creep: How to get the answer you want

NATSEM recently released modelling showing that discretionary tax cuts have more than offset bracket creep since 2005 (that is, taxes are lower than if the income tax scales had been indexed to the CPI or a wage index). The Centre for Independent Studies replied with a study pointing out that you get the opposite conclusion by looking at the period since 2013. Why 2013?

We chose 2013 as the starting year because this is when the last change to tax thresholds occurred

Well, yes. If you pick a period in which there have been no discretionary tax cuts, you will certainly find that discretionary tax cuts have not outweighed bracket creep. As author Michael Potter observes

This shows the importance of the starting year.

There’s no need to check the CIS numbers: given the setup, only one answer is possible.

Correction Michael Potter advises that the CIS study includes the impact of the tax changes in 2012-13, making it possible in principle that these could have offset bracket creep. But the 2012-13 changes weren’t a general tax cut aimed at offsetting bracket creep. They offered small tax cuts for low income earners to offset the very modest impact of the carbon price/tax. These were clawed back by higher marginal rates so that upper income earners (appropriately) bore the full cost of the carbon price. The key point, stated below, is that the Howard-Rudd tax cuts introduced after the 2007 election were so large that they have more than cancelled out all the subsequent bracket creep.

Read More »

Increasing GST: not worth the effort?

The Grattan Institute has just released a report suggesting that the government should get more revenue from the GST, either by broadening the base to include food, health and education (yielding an extra $17 billion) or by raising the rate to 15 per cent (yielding an extra $27 billion). As you’d expect from Grattan, the analysis is sound and careful. As long as you accept the standard framing of the tax reform debate, in terms of the need to shift from direct to indirect taxation, it is reasonably convincing.

Grattan suggest using 30 per cent of the extra revenue to increase welfare payments and 30 per cent in cutting the bottom two tax rates, thereby compensating low income earners. The overview concludes:

Around 40 per cent of the additional revenue from a higher GST would be left over after welfare increases and tax cuts. At least some will need to go to state governments to help them address their looming hospital funding gap, as the price for their support of the change. This would leave a little – but not much – to reduce the Commonwealth’s budget deficit, or to pay for other tax cuts that promote economic growth.

(emphasis added).

Is that enough to sell the package? I can’t imagine the states going along with a deal like this for less than 20 per cent of the total extra revenue, which implies the Feds are left with 20 per cent, somewhere between $3.5 and $5.5 billion. From a political viewpoint, it’s hard to see this being worth the effort for the Turnbull government, especially with no guarantee of success.

As a comparison, the FBT concession for motor vehicles, reinstated by Tony Abbott costs the budget around $1.5 billion. Exemptions for non-profits, which have been comprehensively rorted, cost at least as much. Add in a few ‘rats and mice” concessions, and the Federal government would have as much as it could get, in net terms, from the Grattan package (Getting rid of the non-profit concession would probably require some compensating expenditure, but the same is true of the health and education concessions under the GST.)

That’s before we get to the elephants: superannuation concessions (also supported by the Grattan report), corporate tax avoidance, land tax and higher income taxes for (say) the top 5 per cent of income earners (reflecting elite opinion, the Grattan report suggests cutting these rates). All of these are hard, but not obviously harder than the GST.

So, why is GST reform at the top of the government’s list? The answer is simple enough. The advocates of reform haven’t had a new idea, on taxation or anything else, in 30 years. They didn’t get the GST out of Keating’s Tax Summit in 1984 and they didn’t get the version they wanted from Howard and Costello in 2000. So, the same old idea keeps on coming up.

Populism, Patrimonialism and US politics (crosspost from Crooked Timber)

Nuance is nearly always appealing to academics. For a long time, that was true of my approach to economic issues, particularly including income distribution. When presented with simplistic populist solutions to inequality like “Make the rich pay!”, I was inclined to responses along the lines of “It’s more complicated than that”.

A big problem with “Make the rich pay!” is that with the kind of income distribution that prevailed in the mid-to-late 20th century, any change to income tax that would raise significant revenue would have to apply to the top quintile (20 per cent) of the income distribution. People in the top quintile of the income distribution mostly derive their income from (typically professional or para-professional) employment, don’t think of themselves as rich, and aren’t, in general, seen this way by others. So, the slogan didn’t match the implied policy.

But with the rise of the patrimonial society that’s largely ceased to be the case. The top 1 per cent of the US population now get more than 20 per cent of all pre-tax income, considerably more than the total revenue of the Federal government. Within that group, the top 0.1 per cent have done better than everyone else, and the top 0.01 per cent even better.

So, taxing the 1 per cent more makes sense. I responded a little while ago to a piece trying to argue increasing the top marginal tax rate would make no difference to inequality. And while I was drafting this post, the NY Times came out with an article that reached broadly the same conclusion as mine.

There’s nothing inherently ludicrous in the suggestion that the very rich should pay most or all of the costs of sustaining a system that benefits them so greatly[^1]. And, as in the 1920s, the very rich are different from everyone else. Their wealth is derived primarily from capital, or from control over capital (as business owners or from the financial sector). And, while most of the current cohort of ultra-wealthy did not inherit large fortunes, that’s an inevitable consequence of the fact that there weren’t many large fortunes to inherit until recently. As Piketty demonstrates, a society dominated by large accumulations of wealth will inevitably one in which inheritance, rather than effort, education or talent, determines life outcomes.

Read More »

Would a significant increase in the top (US) marginal income tax rate substantially alter income inequality?


This, you might think, qualifies as another in the series “Short Answers to Silly Questions”. But a Brookings Paper study by William G. Gale, Melissa S. Kearney, and Peter R. Orszag reaches the opposite conclusion. (Hat tip: Harry Clarke).

The study looks at increasing the top marginal tax rate (currently 39.6, applicable to incomes above $400k for singles), with the strongest option being an increase to 50 per cent. The proceeds are assumed to be redistributed to households in the bottom 20 per cent of the income distribution.

The headline finding is that the Gini coefficient is barely changed, as are other popular measures including the 99/50 ratio (the ratio of income at the 99-th percentile to 50-th percentile, that is the median). But the 99/10 ratio and 90/10 ratios change a lot, from 50 and 17 under current law to 37 and 12.5 with the redistribution.

What does this mean? Two things:

(i) As is well known, the Gini coefficient is a lousy measure of income inequality, much more sensitive to the middle of the income distribution than to the tails
(ii) The proposed redistribution would substantially improve the welfare of the poor, with most of the burden being borne by taxpayers in or near the top 0.1 per cent.

It’s obvious, as the authors note, that the 90-50 measure won’t change, since neither group is affected (there’s no simulation of behavioral responses which might have indirect effects). But, since the 99-th percentile income is very close to $400k, there’s very little impact on this group either. But the tax, as modelled, raises a lot of money from the ultra-rich incomes. As a result, distributing the proceeds at the bottom of the distribution raises incomes substantially, which explains the big changes in the 90-10 and 99-10 ratios.

The real lesson to be learned here, one I came to pretty slowly myself is that old-style measures looking at quintiles or even percentiles of the income distribution are no longer very relevant. The real question, in the economy of Capital in the 21st Century is how much should go to the ultra-rich.