Tax churn. How bad is it?

Progressivity of transfers, around 2000: Ratio of benefits received by poorest quintile to benefits received by richest quintile, total population

There’s a new crusade on against tax churning – that’s the state taking with one hand (taxation) and giving back with another (welfare and other benefits like ‘family payments’.)

Peter Saunders of the CIS published three “Issues Analysis” papers on this last year and the memes are well and truly running. Here’s his claim.

At least half of the $175 billion of tax revenue spent on the welfare state last year will probably find its way back to the people who paid the money in. If we could eliminate this churning, it would release $85 billion which could fund spectacular tax cuts without making anyone worse off. We could, for example, raise the tax-free income threshold to $20,000 and combine it with a flat 10% income tax

ALP figures Chris Evans and Craig Emerson have said what a bad thing it is. This is what Senator Chris Evans said. “The cost of this massive recycling of cash is enormous for absolutely no economic gain.”

But a lot of churning is the result of the fact that our tax system targets individual earnings, whilst the payments system is targeting household income. So what Chris Evans is reporting – that a family on a single income of around $50,000 pa with a couple of kids and a spouse is getting back in payments roughly what they pay in tax is no surprise. Of course none of it would be ‘churning’ according to most definitions if it were provided as tax credits rather than payments. But leaving that fact aside, the churning is a direct result of our system supporting families with kids.

I’ll have more to say on this soon, but until I do, have a look at the graph above. Impressive non? The graph is from a very impressive presentation given by Peter Whiteford in the middle of 2005. And this is what Whiteford says “Australia has the lowest level of churning of any country, [in the OECD] at around 5.5% of disposable income”.

None of this goes directly to the merits of the arguments for and against churning. But it’s worth a mention in the debate wouldn’t you think?

This entry was posted in Economics and public policy. Bookmark the permalink.

14 Responses to Tax churn. How bad is it?

  1. Homer Paxton ( where is EP?) says:


    given the Gruen genes are highly correlated with an understanding of economics and the Paxton genes are not does not churning ( whatever the definition) leads to dead weight losses that are very much higher then under the Humphrey’s proposal for example

  2. Neil says:

    Its the single worker who drinks, smokes and drives a car that carries this country

  3. You’ve made (more eloquently, as usual) a point that I was also making about Emerson’s rant. Which raises the question – assuming you want to keep subsidising families and taking money off the childless, how do you do so in a more efficient way?
    On another issue, isn’t the progressivity of transfers in Australia closely related to the “high effective marginal tax rates” that supposedly provide a severe disincentive for welfare recipients from seeing more or higher-paying work?

  4. Good table on churning!

    I suppose the real debate should be about the degree to which we support families with children through the tax system – one whose assumptions are unexamined and perhaps should be.

  5. derrida derider says:

    Is churning inefficient? As economists say, it depends.

    Did you know that most people who send their children to public schools pay tax, and are therefore simultaneously receiving a benefit from government and therefore are churning? And did you know that most people who drive on the public roads pay tax, and this therefore is also churning. And some people even believe that *all* taxpayers get the services of a government-provided defence force (perhaps a view that should be questioned more, IMO).

    The point is, the inefficencies of taxation have to be weighed against the inefficencies of trying to to do strict and immediate user pays. Means-tested family payments are a big source of those EMTRs people whinge about – arguably it would be more efficient, both economically and administratively, to make these payments universal and to claw them back with special family tax scales (this is quite orthogonal to Mark’s point about how much in total we should bribe people who have kids). But this would involve far more churning.

    Contra that, we could reduce churning by a system of very tightly means-tested family payments along with carefully structured child tax rebates to ensure people don’t simultaneously pay income tax and get these benefits, but it is far from obvious that this would be economically or administratively efficient (the interactions when people are working for part of the year only, f’rinstance, would be very tricky. And indirect taxes mean that there would still be churning).

    On top of all that, the measurement of “churning” is very sensitive to the period used. Do we mean people “simultaneously” paying taxes and getting benefits (ie in a single pay period)? Over a year (what about those getting unemployment benefit for the first 3 months of the year and working and paying tax the rest?)? Over a lifetime (we all got the benefit of tax money when we were kids, and most of us will when we’re old)?

  6. Spog says:

    I’d like to add a purely inflammatory comment that churning is the fault of people who think that tax rebates/offsets are better than tax deductions.

    While perhaps not entirely true, there’s more than a grain of veracity in the point.

  7. Spog,

    Sorryk, but I don’t follow the point you’re making, Why is one form of tax break better than another?


    Yes other things equal, targeting will increase effective marginal tax rates as benefits are withdrawn with rising incomes which adds to the tax rate earners are already facing.

    However, when I was looking at this issue a while back, I was surprised to find that for the often shorter income span over hwich benefits are withdrawn, other countries’ EMTRs are often higher than Oz – because they’ve not given as much thought to the design of withdrawal regimes.

  8. Spog says:

    Sorry about that Nicholas, it was a red-wine influenced posting. I had been thinking that some of the direct (or “cash”) payments that are now made started out life as tax deductions.

    The life-cycle of these things seems to have been:

    born as a tax deduction (eg, for a dependent spouse);

    transformed into a tax rebate (eg, dependent spouse rebate/tax offset), in response to [bizarre] concerns that the amount of the benefit obtained as a deduction varies with the tax bracket you are in;

    transformed into a cash payment (eg, home child care allowance, which became basic parenting allowance, which became family tax benefit part B).

    In the cash payment guise, it becomes criticised as either an example of pointless churning, or as middle class/high income earners welfare. Hence my point that issues with churning stem from misadventures with tax deductions. However, I’ve already apologised for the red wine, so I won’t go any further into the mire.

  9. Whiteford’s paper is indeed very useful. One of the points he makes is that “In virtually all OECD countries, the middle 60% of households receive between 50 and 65% of all transfers” with Australia at 56%, so “Australia has roughly the same share of middle class welfare as most other OECD countries.” Churning overall is less than elsewhere because we more successfully exclude the top 20% from getting many benefits (in contributory social security systems in Europe, everyone who pays into the system is entitled to take out of it).
    My point about our high level of middle class welfare is that middle income people should be paying less tax and relying more on their own efforts, rather than being driven into dependency on government by high taxation and then absorbing the lion’s share of government cash benefits and services to make up for it.
    We need to reduce both simultaneous churning (a lot of which is down to the family payments system) and lifetime churning (which could be reduced by encouraging saving). My third paper on churning ( has some ideas how we might start to do this.
    Why should we do it? Because taking tax off people and then giving it back to them is (1) costly to administer (2) the major cause of high EMTRs which produce work disincentives (3) disempowering and destructive of the spirit of self-reliance (4) an invitation to pressure group rent seeking (5) socially destructive (because it crowds out voluntarism and encourages a culture of civic irresponsibility).
    We should welcome Labor’s initiative in putting this on the agenda.

  10. Peter,

    Thanks for your response. I’ve read your papers and my own bias is the same as yours – to encourage self-provision where possible. But Whiteford’s paper seems extremely well informed and it raises a range of points – such as the sentence I quoted on how Australia is the lowest churning OECD country – which I would have thought would have been important to inform your readers of when they read you on this subject.

    Also, Whiteford’s comments on churning runs for seven pages including over a page of specific responses to your suggestions. He’s fairly sceptical of your proposals without rejecting them out of hand.

    (For instance just trying to think it through intuitively, it seems to me that using the tax free threshold is a very blunt instrument because it is so indiscriminate. This means that it gives away large amounts of revenue and so undermines our capacity to cross subsidse those who are in poverty more than temporarily and families with children. (I’ve take it from comments you’ve made in various publications that you’re broadly in support of these goals, even if you want to scale them back to the extent you can bring about self-provision.)

    Can you tell me if there’s a detailed response to Whiteford’s comments.

  11. derrida derider says:

    Nick has brought out more clearly than me the problems in Peter’s position on self-provision. Basically, the laws of arithmetic are inescapable.

    If you want to tax the better-off to give money only to the “truly poor” then you cannot avoid giving the latter an incentive to stay truly poor. If you address this by also giving money to the not-all-that-poor then you’re straight into issues of middle-class welfare. Not to mention that you’ll be worsening the horrors of taxation for the better off; perhaps the real motivation behind a lot of right-wing commentary on these issues (Peter, I’m not accusing you personally of this last but it certainly figures in the calculations of many of your ideological friends).

    It’s not at all clear that avoiding all “middle class welfare” and “churning” is economically and administratively efficient. A pure BI/FT scheme is the ultimate in middle class welfare and churning, yet much of the optimal tax literature tells you its the most economically efficient method of redistribution, and it’s certainly the most adminstratively efficient.

    The other way to tackle all this is to muck around with non-financial incentives. That is, beat people about the head so hard that they’ll be desperate to work even if its not financially worth it. Apart from some obvious social and equity implications which well-fed commentators seem far too relaxed and comfortable with, there are obvious practical limits to this approach.

  12. derrida derider says:

    Oh, and BTW Robert Merkel’s rant (linked to from his comment above) is pretty good for someone with no background in the issue – its worth following the link.

  13. spog says:

    The inflexible laws of arithmetic may indeed mean we have to pay something to people on higher incomes, but to some degree the amount required is conditioned by societal expectations.

    If there was no welfare system, and people were forced to get insufficient incomes topped up from brothers or sisters, etc, what would the prevailing view be about the appropriate EMTR settings to encourage worforce participation? For example, if a person needed $300 a week to live on, but only had a part-time job paying $200, they’d need to get $100 a week from their family. If that person then had a pay rise to $250 a week, their top-up requirement would now be $50. I suspect that in a system where the person paying the welfare was clearly identifiable (eg, they were your brother), there would be little qualms about saying I now only need $50 to top me up to the necessary minimum income. In other words, in that context a 100% withdrawal of income support would be acceptable.

    In our system, the payers are far removed from the the recipients. We also live in a market economy where the pursuit of a profit results in a profit and loss calculation for most actions. Does this inevitably inculcate a view that the withdrawal of welfare payments must be at rates which still allow profits to accrue to the recipient? Perhaps you can’t condition a population to a market view of the world without also accepting that even the poorest will do market-style calculations of gains that will or won’t flow from their actions.

    We used to have 100% withdrawal rates on income support payments. Sure, this was back in the days when getting a job meant getting full-time work, so there was no concept of having to accommodate part-time workers, but it was also a time when receipt of the dole was a massively stigmatising prospect. I think times, and the extent of our market orientation, have changed since then. Middle class welfare may be a necessary adjunct to our style of society, partly because of the laws of arithmetic alluded to by Derrida, but also because we have been conditioning people to think profit first.

  14. Pingback: How Much Does a Basic Income Cost? | Green Agenda

Leave a Reply

Your email address will not be published. Required fields are marked *

Notify me of followup comments via e-mail. You can also subscribe without commenting.