Progressivity is not the same as redistribution

Peter Whiteford is one of my favourite commenters. He rarely joins a thread without adding useful data or some telling insight. On Monday he showed up on Matt Yglesias’ blog to explain the difference between progressivity and redistribution in the tax system.

The debate was over a table Scott Hodge posted on the Tax Policy Blog. According to Hodge "the U.S. has the most progressive income tax system among industrialized nations." Yglesias argued that this was misleading. Here’s part of Whiteford’s reply:

… progressivity is not the same as redistribution. Progressivity measures how the distribution of the tax burden is shared, while redistribution measures how much the tax system reduces inequality. Redistribution is influenced both by the progressivity of taxes and the level of taxes collected.

In fact, the US system of direct taxes actually reduces inequality more than any other country as well. But overall, the USA reduces inequality a lot less than most other countries, because the other thing that you need to take into account is what taxes get spent on.

Now the US system of social security and cash benefits reduces inequality by less than any other OECD country except Korea. The US social security system is marginally less progressive then the OECD average, but the level of spending is very low – only Mexico and Korea spend less in the OECD.

So while the US tax system is progressive and reduces inequality, the US welfare state is much less effective at reducing inequality. And because the US has a very unequal distribution of income from capital and a much wider wage distribution than many other OECD countries, it ends up as a relatively unequal country after taxes and benefits.

Of course Whiteford is well placed to explain the data. As he wrote in his comment: "I am the person who wrote the chapter in the OECD report that is the basis of these figures."

Greg Mankiw reposts Whiteford’s comment in full. And there’s more discussion from Scott Sumner at The Money Illusion. The table is over the fold:

This entry was posted in Uncategorized. Bookmark the permalink.

31 Responses to Progressivity is not the same as redistribution

  1. Martin C. Jones says:

    Peter Whiteford has a piece in the Economic Record (Vol 86 Issue 275) on this very topic. It really is very good.

  2. Incurious and Unread (aka Dave) says:

    This is the kind of thing that gives economists a bad name. Redefine the meaning of common words. Then use those new meanings to come up with startling, counter-intuitive results.

    It works so well that even economists can surprise themselves.

  3. Patrick says:

    Au contraire, Dave, I would have thought this was the kind of thing that gives economists a good name!

    I find it very useful to have distinctly defined terms in a debate, not to mention the empirical data to support them.

  4. Peter Whiteford says:

    Thanks Don (your cheque is in the mail!).

    For people who can’t access the Economic Record there is a much longer version of the paper at which has all the gruesome details.

    On Dave’s point, this is not an attempt to redefine the meaning of common words. Virtually all of the previous studies of tax progressivity use the same approach.

    High income groups in Europe will generally pay a higher share of their total income in taxes than do high income groups in the USA or Australia. The point is that as well as high income groups paying higher taxes in Europe, low income people pay much higher taxes, as well. Progressivity measures the difference between the taxes paid by high and low income groups.

  5. wizofaus says:

    Wouldn’t that mean Europe generally has more churn than the US or Australia?
    If so, I wonder if the benefit of this extra churn is the sense that “everybody puts in and everybody gets back”, and if so, is it one that could reasonably be said to be worth the cost?

  6. Peter Whiteford says:


    Absolutely correct. Australia has the second lowest level of churning in the OECD (expressed as a % of disposable income). Korea is the only OECD country with lower churning, because you don’t have much churning when you don’t have much of a welfare state (their social security and health spending is about one-third of Australia’s.) Churning in most european countries is between 2 times and 5 times as high as in Australia. See Table 3.6 in the link above.

    It is correct that much of what gets labelled as churning is redistribution across the lifecycle which can also reduce inequality. However, there are also problems with estimating how much inequality really gets reduced because these are point in time data, and the amount of redistribution between rich and poor at a point in time gets mixed up with redistribution across people’s lifetime (which is a complicated issue).

  7. Incurious and Unread (aka Dave) says:

    Peter and Patrick,

    Suppose Country X levies an income tax of just 1% on the top income earners and nothing on anybody else. The tax revenue is used to pay for law enforcement only, with no social security etc.

    Suppose Country Y levies a flat income tax rate of 30% on everyone and uses the revenue (inter alia) to provide welfare payments to those on low income and support free healthcare, education etc.

    Economists would say that country X has the more “progressive tax system”.

    Nobody else would.

    That is my point.

  8. Peter Whiteford says:


    I don’t normally rely on Wikepedia but for example” A progressive tax is a tax by which the tax rate increases as the taxable base amount increases.[1][2][3][4][5] “Progressive” describes a distribution effect on income or expenditure, referring to the way the rate progresses from low to high, where the average tax rate is less than the marginal tax rate.[6][7] It can be applied to individual taxes or to a tax system as a whole; a year, multi-year, or lifetime. Progressive taxes attempt to reduce the tax incidence of people with a lower ability-to-pay, as they shift the incidence increasingly to those with a higher ability-to-pay.”

    The online access to Oxford English dictionary defines progrssive as follows: “of taxation or a tax) increasing as a proportion of the sum taxed as that sum increases”.

    This is also the way that is standard in economic analysis of progressivity.

    So if we use this definition then Country X does indeed have a more progressive system than country Y.

    But as you point out – and I also have pointed out in boring detail – the overall tax and transfer system in Country Y is much more resistributive than in country x.

    Which is why I made a distinction between the two terms. I think that we are both in agreement on what actually reduces inequality, but I am not redefining the terms.

  9. Incurious and Unread (aka Dave) says:


    Sorry, I wasn’t suggesting that anybody was redefining the technical economic terms being used or using them in the wrong way.

    I was rather suggesting that these terms – and hence the meaning of the findings – could easily be misinterpreted by a lay audience not familiar with the technical meanings.

    Certain high-profile economists, with widely-read blogs and a particular economic agenda to push, may use this potential misinterpretation to mislead the public at large.

    It is this disingenuity and deception which I am suggesting can give economists a bad name.

  10. Fyodor says:

    Dave [are you “Haiku” Dave, by any chance?], I think the only confusion, disingenuity and deception there is amongst people with notions of ideological (i.e. socialist) purity around the word “progressive”.

    Peter, interesting work and great to have you on-line here at Troppo. How did you treat the typical forms of “middle-class welfare” in Australia, i.e. the subsidies to private education, private health insurance, child care and the “tax expenditure” on negative gearing? Are these included in your calculations on fiscal churn?

  11. Peter Whiteford says:

    But I was also using the standard definition in the dictionary.

  12. Incurious and Unread (aka Dave) says:


    I hope you do not think that I am criticising you. Quite the reverse. You are taking the trouble to explain (in your comments on the Yglesias blog and here) that “progressive tax system” does not imply a high level of redistribution, as the casual reader may have thought.

    Rather, my criticism is of Mankiw for providing the blog headline without any attempt to explain precisely what it means and for linking to a post at the “Tax Foundation” website headlined:

    No Country Leans on Upper-Income Households as Much as U.S.

    As far as I know, “leans on” is not a technical term in taxation theory. If my “Country X” existed (see my comment @7 upthread), the upper-income households – paying a 1% tax rate – would take the title of “most leant on”.

    I mean, really.

    Your research is being wilfully misused and misrepresented. I would have thought that you would be more upset than I am.

  13. Peter Whiteford says:


    Sorry if I come across as grumpy. I totally agree that “leant on” is a misleading term – and if either the Tax foundation or Greg Mankiw allowed comments, I would have commented as such. Obviously, however. I think that progressivity is the correct word.

    Fyodor, the calculations of things like churning given in the paper linked to only include cash benefits and direct taxes, since household income surveys don’t collect information on services. Private education subsidies for example are not directed to households, but to schools, and while they can be modelled they are not routinely included in income surveys.

    Tax subsidies for private health insurance or negative gearing will be measured as reducing churning, which is why I think that churning is not a very useful concept.

  14. Incurious and Unread (aka Dave) says:


    I can live with “progressive”. Its technical meaning is broadly aligned with its “lay” meaning associated with the promotion of social and economic equality.

    “Tax System” I have a problem with though. Your link @4 is to a presentation you did for the Henry Tax Review. Have you paused to think about the title of that review: “Australia’s Future Tax System“. What does “Tax System” encompass in this context?

    The Australia’s Future Tax System Review was established by the Rudd Government in 2008 to examine Australia’s tax and transfer system

    Not the fault of economists, I know. But it is no wonder that people are easily confused and misled.

  15. observa says:

    Let’s take Australia A as of right now and then compare it with Australia B in say 2050 where Australia B still has exactly the same redistribution in real terms but, except that now Australia B apart from the anomolies of excise on booze and fags, now collects all previous forms of taxation via CO2E taxing. Which is the more progressive and/or redistributive tax system as a result?

  16. observa says:

    You’re kinda walking in the footsteps of giants like John Rawls, except like asking the Irishman directions to Dublin he reckons you wouldn’t want to start from here. Rawls likewise was locked into his own dumb original position but unlike him we have a certain advantage of experience and much later hindsight. In that regard you’ll note Australia B just nailed this thorny problem among others-

  17. observa says:

    Yes it’s a much tougher question when you’re so wedded to income tax progressivity and equity, but Australia B does address the notion of intergenerational equity that so many have been banging on about of late. So let’s ask a simpler question, seeing that now Julia and Co reckon we should be relying on more resource taxes as well as ‘other’ tax cuts not being ruled out. Australia BB now has complete reliance on a mix of CO2E taxing and resource taxing only, all other things being equal. Does that make Australia BB more or less progressive/equitable than simply Australia B?

  18. Peter Whiteford says:


    I’m not sure exactly what tax system design you have in mind. I get the impression that your hypothetical scenario compares our current system with one where all the revenue is collected through carbon taxes – but correct me if I’m wrong. Depending on how you did this, it seems likely to me that such a tax system would be less progressive than the current system.

    Now you could potentially offset this through changes to benefits, and end up with the same distribution of income as currently have (assuming nothing else has changed). In this case the overall tax-benefit system is not less progressive than currently. Put another way, we need to know what the overall impact of the taxing and spending is, but we still want to be able to characterise accurately what is happening in the different components.

    But I would also argue that progressivity is not the only criterion by which we should judge the design of the tax system or the benefit system. For example, it is clear that taxes on tobacco are regressive. But lung cancer and smoking-induced heart failure are also regressive. We live with a regressive tax because we think that the broader social results are better.

    Having said this, it doesn’t make sense to me to say that tobacco taxes are not regressive. They are – it is just that society has other goals which it judges are more important than progressivity.

    I would argue that you need to look at the overall system of taxes and benefits to assess what is desirable, but it is still more accurate to say that a particular tax or benefit is progressive or regressive. We need to be clear about both the overall effects and the effects of each component.

    It is clear that environmental and economic sustainability is pretty fundamental. My view, however, is that it doesn’t help to use the wrong words or concepts to describe different issues. Terms like progressivity and redistribution have specific meanings and we should use these terms precisely rather than try to encompass different concerns within the one label.

  19. Incurious and Unread (aka Dave) says:

    observa @15,

    I think that aim would be problematic. Australian annual greenhouse gas emissions are around 400m tonnes of CO2 equivalent. Its tax revenue is around $400bn. So, the carbon price would need to be around $1000/tonne to recover this revenue from carbon taxes alone.

    AIUI, the target for 2050 is a 90% reduction in carbon emissions, meaning the price would need to rise to $10,000/tonne.

    At these prices, there would not be a whole lot of carbon emissions left to tax.

  20. wizofaus says:

    I’m curious actually, have anyone tried to come up with what our tax-and-benefit system would be like if *all* taxes collected where based on, say, environmental externalities such as CO2 emissions, but where we wished to preserve more or less the same effective net income distribution as we have now? It seems to me it would only be possible by having the tax rates very high so that, say, someone currently earning 200K a year who pays roughly 65K in income tax would then be paying no income tax but paying 65K *more* for their goods and services – which would imply environmental taxes are at least doubling the cost of these, but then somebody earning 20K a year (who currently pays ~2K income tax), assuming they’re currently spending even only 10K a year on goods and services subject to such taxes, would need to receive benefits in the order of 8K in order to have the same purchasing power they do currently. But it still seems to me that those earning over $1M, unless they spend a very significant fraction of their income on goods and services particularly vulnerable to environmental taxes, would be in a considerably better position than they are currently (assuming they do actually pay the expected income tax!), and hence a larger share of the tax burden would be falling on those less able to pay it.

  21. Incurious and Unread (aka Dave) says:


    Another way of framing your question is whether a person’s environmental footprint is proportionate, “super-proportionate” or “sub-proportionate” to their income.

    Booze, fags and pokies may be the vices of the working class, but the rich have their own vices: large cars, houses, gardens, swimming pools etc. I don’t know the answer to your question, but it is far from obvious to me that a broad tax across all of these environmental “bads” would be regressive.

    Another issue is whether there really enough “bads” from which to raise tax revenue. Part of the problem is that many potentially lucrative bads (eg illegal drugs) are prohibited and hence not available to be taxed. But even with widespread legalisation, there may simply not be enough bads to fund government.

  22. wizofaus says:

    Well it doesn’t seem entirely fanciful to think that a significant benefit of only taxing ‘bads’ is that the cost of government may well be significantly reduced.

  23. Patrick says:

    I’m happy to tax only bads if we only tackle bads with a direct external harm, or, if we include indirect causationally difficult harm, we include expression of socialist thought. $150 a socialist-inclined word, as interpreted by me.

    Actually I don’t really know what wizofaus means. Do you mean reduce the amount of government inputs and thus necessarily the outputs, or do you mean improve the efficiency of existing government (in which case, how do you expect that to happen based on taxing ‘bads’? Tax collection alone is not such a big expense of government…

  24. wizofaus says:

    The point of taxing ‘bads’ is to discourage our tendency to engage in behaviours that require somebody else to pick up the cost later. That ‘somebody else’ tends to be the government. E.g. tax fatty foots and you end up with less obesity, hence less needs spending on health care.
    But yes, I’d also expect some savings from a radical simplification of the tax system.

  25. observa says:

    It would appear post Copenhagen that we’re a fair way from that hypothetical $1000/T CO2E tax that may be the ultimate goal for many nowadays. That assumes a C&T scheme built upon some fresh faced kid come a knocking to change some light bulbs and shower heads and doing some quick average calc of CO2 thereby ‘saved’, to be traded by the Morgan Sachses of this world, with all its other jurisdictional rorting and negative externalities (putting the worlds food in our tanks and knocking over Asian rainforest for biodiesel spring readily to mind here)has you thinking like Abbott and Co, that that’s a load of crap. However, to the extent that we reduce all the other forms of taxation, in particular income tax, that allows us to implement higher CO2E taxing and resource taxing generally. In fact with the MRRT the Govt is touting the ability to reduce company tax, along with income tax cuts being on the table with a shift to carbon taxing.

    Wizofaus ponders-
    I’m curious actually, have anyone tried to come up with what our tax-and-benefit system would be like if *all* taxes collected where based on, say, environmental externalities…
    Well yes I’ve given it a great deal of thought and believe it’s entirely possible to conceive of such a blueprint, if not the actual final level of the respective mix of taxation within that blueprint. Whilst I might proffer a logical overall blueprint, as you’ll see I could hardly be expected to know the emphasis on the final mix. We would need to leave that to evolution by the political system. Nevertheless that’s why I challenged you here with some hypothetical questions and also ask you to look carefully at the markers about you to see where that blueprint lies.

    We’ve seen the argument for CO2E taxing and also an argument for resource taxing largely on the basis of ‘make the big polluters pay’ and ‘they are using our collective resources’. Well isn’t that simply true of all of us and if so why not a level playing field here? There is no ‘us and them’ but one great melange of us all utilising our natural environment for our needs and wants to varying degrees, with very little countervailing market power to its ultimate depletion/destruction. Just ask John Walmesley with his Earth Sanctuaries Ltd about the absolute truth of that. To understand the move to greater resource taxing was to comprehend the critical importance of land use in that equation. Henry George had much to say about that but he was a captive of his times. A child of endless frontiers, rather than 1969 spawning Spaceship Earth and our satellites subsequently overseeing it all and changing the generational perspective.(in fact the AGW apocalypse is largely all about that) You have to tax land use as a resource, but not on its value, for to do so would encourage its commercial conversion/utilisation as we’ve all seen.

  26. observa says:

    For me progressivity and equity are rather outmoded terms now that the environment and associated intergenerational equity have seriously entered the social conciousness and the markers are everywhere-
    Like most left/greenies Minister Caica gets the bit about the pressing need resource taxing but not the obvious tradeoffs necessary to seriously implement such a regime. Something Frank Sartor was warning Labor about last night on Lateline after the caning NSW Labor received.

    So to help left/greens out here I’ll finish outlining a serious market based third way through the mire for them. Something we all so desperately need after left/green Govts everywhere have been so busy picking losers as Sartor noted so poignantly. On top of rollbacks to green loans/green inspectors, solar feed-in tarriffs and RECs schemes, here in Rannistan the Govt has just junked the ‘Water for Good’ scheme whereby subsidies were handed out for plumbed rainwater tanks and dual flush cisterns. This after previous junking of water efficient washing amchine subsidies. Perhaps some can still take solace in Council worm farm, compost bin and garden mulch subsidies.

    But I digress and we were up to a mix of CO2E tax, resource taxing including land use. Now land use needs to be taxed at zero for natural environment up to a maximum for building/concrete/bitumen cover. We can easily gauge that like Councils do now with Google earth. That is fundamentally to price the use of the natural environment with biodiversity in mind. Essentially the natural environment needs true countervailing market power to it’s destruction (those McMansions and the march of tuscan boxes) But it needs more than that. It needs the market power to actively push back and create more natural environment a la Walmesleys Earth Sanctuaries experiment. A noble experiment but one doomed to failure due to its reliance on altruistic donations rather than a reasoned constitutional marketplace set by our taxation system. To that end we’d only need to add an annual net wealth tax for equity at the very top end of the wealth/income stream, essentially for those who have enjoyed the maximum benefit of our social/communal system, with an important exemption. All wealth held as natural environment (ie in trust for us all and future generations0 is exempt and any expenditure on rehabilitating or reinstating same, earns a franking credit against any ANWT liability.

    That’s a simple, level playing field, green constitutional marketplace that the rest of the world will have to follow more or less(can you see why?)and one which will hopefully banish the human hubris and folly of picking so many losers to date, not to mention the regulatory quagmire surrounding it all. Adam Smith long ago appreciated the superior, continuous spectrum of price over lumpy, coarse, quantitative and legislative controls and yet it is a salutary lesson lost on so many nowadays. I claim that blueprint is unbeatable in addressing true environmental and equity concerns at least cost with maximum benefits and outcomes and challenge any of you to better it. You throw up a problem and I’ll show you how that blueprint solves it and at best all you’ll be arguing about is over the emphasis on particular tax mechanisms. Here’s one out of left field it seriously addresses, after all the failures of your current constitutional marketplace to date-
    Can you see how?

  27. observa says:

    While I just put Peter Whiteford out of a job writing endless reports on those vexed effective marginal tax rates and the interaction of income tax scales with pension/benefit withdrawal rates, unfortunately I’ve also put all the xenophobic folk at the FIRB on the dole queue as well. No matter there’s opportunity aplenty out there with no tax impediment whatsoever on human ingenuity, entrepreneurship and good old fashioned sweat anymore. We’ll need plenty of that as the price of labour falls relative to that more costly CO2E capital and whole new environmentally incentivised industry springs into action with no more frictional hurdles like stamp duty and capital gains tax triggers to prevent the free marshalling of environmentally aware capital to the best use. Look out Gaia here we come eh? A lot more economic circumspection about ramping up the most expensive housing in the world too I’ll wager. Then there’s all that failed intervention with the first inhabitants, with little but their physical labour to sell, although the well meaning interventionists reckon they have the closest affinity to Gaia among us all. Look out John Walmesley eh? You’re going to have some serious competition for all that countervailing market power and ‘investment’ in Gaia in future.

    Meanwhile, back in left/green quant nirvana (well the Saudi Arabia of uranium perhaps), the usual suspects are busilyy investigating whether Garbags, sandwich bags or Gladwrap, etc should be next cab off the rank to be added to their supermarket plastic shopping bag ban. I had a dream…

  28. Peter Whiteford says:


    I am not so easily put out of a job. I am planning to have a new paper on effective marginal tax rates coming out around the middle of the year ….

  29. observa says:

    Ah well Peter, I almost had you working on the subtleties and nuances of that ANWT and how you’d design it equitably for the life cycle and the like. Either that or for one of those many multinationals setting up head offices here with no company tax, just paying their carbon and resource footprints, if the ROW were silly enough not to to emulate us ;)

  30. observa says:

    And speaking of subtleties and nuances, here’s one for you to ponder Peter. John Quiggin raised the issue once as I recall about certain ephemeral wealth being ‘subsumed’ or ‘enjoyed’ temporarily a la sending the kids to private schools versus the public domain and the higher benefits obtained. Now by the unanimous comments I take it we’re all agreed on this blindingly obvious new third way constitutional marketplace of ours, whereby the private school is paying its fair share of CO2E tax, resource taxes and land tax based on its building/bitumen/etc cover of Gaia. In that regard metro area land holdings may also have a percentage factor added for the paved footpaths and roads their owners enjoy vis a vis a rural holding or school. So far so good and all the teaching staff have their gross incomes in their weekly hands as do the parents, with some paying the ANWT deeming rate on any wealth not held as natural environment at the time. But who ‘owns’ the wealth of the fine private school built on the contributions of the old school tie and perhaps the odd BER largesse? Should all such amorphous wealth not in public hands for full public access be allocated for ANWT purposes. That was where Quiggin was coming from. Should nominal shares in say the church school be held/allocated among the church officials, parents, old scholars, investors or whatever for ANWT purposes? Quiggin would no doubt say so if I read him correctly.

  31. Pingback: Club Troppo » Minor Blog Wars – my part in their genesis

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.