Sometime commenter Spog sent me the diagram and the commentary below on the question of churning. He’s produced an excellent diagram illustrating the incidence of churning. It seems to work approximately as one would hope, to target assistance where it’s most needed – subject to the constraint of withdrawal rates that keep EMTRs (always a problem with targeting) tolerable.
Spog’s commentary is as follows.
The chart shows the range of private income over which some selected household types receive government payments. Each bar represents the payment and its thickness gives some indication of the relative amounts paid in each case. Each bar is colour coded into the amount of the payment that is simply churned back as tax (the green component) and the resulting net payment (the yellow component). This is meant as an approximation of point-in-time churning in the tax-transfer system, not “lifetime” churning.
Some churning occurs with all the depicted households, although at the scale used here, this is a little hard to see for a single person. Churning is most apparent in the households with children. The two single income examples (the single parent and the single income couple) have payments that extend off the scale – the oft criticised family tax benefit part B (FTB B). The extension of child related payments to relatively high incomes (unlimited in the single income households) illustrates the middle-class welfare issue.
Perhaps the most important thing the chart shows is not that middle-class welfare and churning are rampant, but that, with the possible exception of the single parent results, the tax-transfer system produces net benefit outcomes that are arguably quite reasonable on a comparative basis. The Australian system is intended to have some regard to household composition, both in terms of the amount of cash assistance provided where there is no private income, and the point at which differently constituted households become net taxpayers. Whether the churning approach is the best way to achieve this makes for entertaining debate.
Compare too, the relative positions of single income and dual income couples, with and without children. At first glance, single income couples without children do not have the open-ended cash payment of those with children FTB B. However, they do benefit from the spouse tax offset, invisible here because it is not a cash-based entitlement. It is open ended, but its invisibility seems to have enabled it to escape the middle-class welfare criticisms regularly directed at FTB B.
I’m beginning to change my mind on the churning issue. Maybet we should try and eliminate this sort of churning, but not because it’s inefficient (as I’ve argued elsewhere, it often isn’t compared with alternatives that meet the same goals). Rather, spog’s right – a tax break is not seen as “welfare” and so retains support even where it’s doing the same thing as a cash payment. All those right-wingers should, if they’re Machiavellian, be pushing for even more churning on these grounds in order to undermine support for redistributon.
The other traditional argument for cash payments – that it can be directed to the mum (who’ll buy clothes for the kids) rather than the dad (who’ll just pour it down his throat in the pub) is not as strong as it once was, and in fact can be turned around – paying it to the earner makes mum want to be an earner.
But once you go down this path, you start to get strong arguments for a radical redesign of the income tax/welfare systems. You’d have very stringent means tests with incentives ameliorated by a large EITC – along the lines pushed by Dixon, Foster and Gallagher in the early 80s and David Ingles in the early 90s.