Progressive taxes vs progressive policies
by Jonty Olliff-Cooper
Hiking up taxes that take up more of poor people’s income than richer people’s might actually be a more socially just way of running the tax system - that was the controversial claim of some at a seminar on the future of taxation Demos held in the heart of the City of London last week.
The idea is pretty hard to agree with at first. Indirect taxes, like VAT and alcohol duty, contribute to the unfair burden of tax on the poorest in Britain. The major reason is that although direct taxes are often stepped up as you move up through the income brackets (the technical definition of ‘progressive’ used by economists), indirect taxes are not. You pay the same for a packet of cigarettes regardless of your income.
We tend to think of our tax system as skewed to hit the rich harder than the poor. If you are lucky enough to be in the 40 per cent income tax band, you perhaps feel that once you add in national insurance and VAT, the state is getting on for taking about half of your income. In fact, if you earn that much, at present you are getting away lightly. Analysis by Compass suggests that the total pain of tax is pretty evenly spread across all income deciles at about 36 per cent of pre-tax income, with the exception of one group who pay much more of what they have: the poorest tenth.
Source: Byrne and Ruane, S (2008), The UK Tax Burden: Can Labour be called the “Party of Fairness”?, Compass Think piece 40, Compass.
This seems a clear injustice. The poorest people should not pay the highest proportion of their income as tax. Several groups on the Left have used this sort of analysis to call for higher rates for high earners.
But if your aim is to tackle poverty or inequality, setting tax policy is not as straightforward as that. One surprising conclusion held by some attendees was to challenge the notion that the most progressive (in its broader meaning as ‘pro-poor’) end result is to develop a progressive (as in stepped) taxation system.
This is because tax rates are not the only way to tackle inequality. Although direct taxes put the burden on those with the broadest shoulders, in some cases direct interventions such as education spending can be a more effective route to building a fair society. By this logic, it makes sense to maximise the tax take first – by whatever means - then redistribute it later via benefits or public spending. Since indirect taxes are harder to avoid, easier to impose and cheaper to administer, they might be a politician’s policy of choice ahead of weathering political resistance to headline income tax rises.
The question asked by those at the Progressive Policy Forum was: is raising more money, regardless of its origin, more important than who pays for it? If more money for services like SureStart is raised from indirect taxes policymakers need to think carefully about whether their core aim is to generate maximum revenue (wherever that might come from) or redistribute wealth and services by targeting taxes at the rich, even though this might not generate as much cash.
So paradoxically, indirect taxes might just be more progressive overall in some circumstances. This is something for progressive conservatives to mull on: the implication is that proponents of a fairer society should not dismiss regressive taxation out of hand as a tool to achieve social justice.
Common Man
And this from someone who is supposed to be one of the fresh new faces within the Conservative party? I suppose pre-established views about what the party is and who it stands for aren't going to be swayed much in the near future. There is nothing progressive about taxing everyone similarly, regardless of income, and then redistributing the resulting procedes. Just ask the former Soviet Union.
Anonymous economist
The lack of basic economic understanding in this post is cringe-worthy.
The author's argument appears to run as follows:
Assumption 1: Spending can be progressive as well as taxation.
Assumption 2: A consumption tax will achieve a higher revenue than a (progressive) income tax.
Conclusion: Therefore, the best way to achieve a reduction in inequality is to use a consumption tax accompanied by progressive spending.
The first assumption is entirely correct. Here, the author makes a fair (if rather dull and well-established) point: inequality can be tackled through spending as well as through taxation.
It is worth noting here that whilst spending can be pro-poor, it is not necessarily so.
However the second assumption is embarrassingly flawed. It is simply incorrect to assume that a consumption tax will achieve higher revenues than a (progressive) income tax. Either can achieve the higher revenues, depending crucially on the rates of tax that are chosen. Additionally, many economic commentators believe income tax to be more efficient than a consumption tax.
Hence, the argument that a consumption tax may be optimal because it achieves a higher revenue simply does not wash.
What's more, the author appears to assume that a consumption tax is neutral (as opposed to progressive or regressive) because the rate is flat. This is simply not the case in practice. Because the poor spend a higher proportion of their earnings than the rich do, a flat rate consumption tax means the poor pay a higher proportion of their earnings in tax.
Kaveh
Anonymous Economist: Your post is pretty overblown and mostly besides the point.
Contrary to the suggestion in your final paragraph, it seems clear Jonty knows consumption taxes are regressive. In his second paragraph, he states: "Indirect taxes, like VAT and alcohol duty, contribute to the unfair burden of tax on the poorest in Britain." His posting of the graph and discussion of it make it clear he's aware that indirect taxes have a regressive effect.
He also didn't assume consumption taxes always raise more revenue than income taxes, regardless of tax rates (that would obviously be absurd). His thought experiment does rely on them being more efficient, which may be debatable but is certainly not "embarrassingly flawed". It's not an assertion, it's an assumption for an interesting thought experiment.
The point is *if* regressive consumption taxes raise revenue less wastefully and more efficiently (i.e. less cost and unintended consequences per pound raised) than progressive income taxes, they should not be opposed by progressives, *as long as* the revenue is spent in progressive ways. The caveats are important.
It's a thought experiment. The aim is to make progressives think more carefully about something they might have instinctively dismissed. It seems self-evident to me that if consumption taxes are more efficient than income taxes, they shouldn't be dismissed out-of-hand by progressives.
You could be more constructive by posting some of the reasons why "many economic commentators believe income tax to be more efficient than a consumption tax", because that is the crux of the issue. I am not sure which is more efficient, but Jonty's arguments ("indirect taxes are harder to avoid, easier to impose and cheaper to administer") seem common-sensical to me. I'd like to hear your arguments as to why income taxes are more efficient than consumption taxes so we can actually get a balanced and informative debate.
Anonymous Economist
Haha - at least it's sparked a bit of debate!
I accept your point on the author acknowledging that consumption taxes are regressive (should have concentrated through to the end!) although I do think the 'technical definition' of a progressive tax is weak - a progressive tax is one that demands a larger proportion of income the higher the income is. A stepped up income tax does achieve progressivity, but it is a long way from defining it.
More importantly, even re-reading this article, and picking up a bit more on the tentative tone and use of the word 'if', it still reads to me as though the author does assume that a consumption tax is a better way to go about raising that income. The main point in my previous post is that this simply can't be taken for granted. Either one can raise more income depending on the rates, and issues of efficiency are hotly debated.
If you asses which is preferable against an equity criterion, then (unless you get into some of the less common definitions of equity) by and large an income tax will win the debate simply because it provides more flexibility to adjust with income than a consumption tax.
On efficiency, I think there are good arguments on both sides (I'll let you google them rather than listing them here, since I don't feel strongly either way) - but the key point to debate is the arguable negative effect that an income tax has on savings and investment. Administrative costs etc are minor in comparison and I don't know which is greater, but it doesn't seem obvious that the costs are higher for an income tax.
Avoidance on income tax is phenomenally high and I don't doubt higher than on consumption tax currently - but it strikes me that this is not because it is easier to avoid but because there is far more incentive to avoid it. If you put consumption taxes up to the rate that would be necessary to eliminate income tax, think what would happen to the black market of imported goods!
But talking of avoidance here is also a bit of a red-herring - because we are talking about setting optimal policy, and therefore any tax rates is at our disposal, avoidance actually just changes the rates chosen and who pays, not revenues raised. So it's more an equity argument than an efficiency one in any case.
I think the argument as you state it - highlighting the 'ifs' and adding the caveat 'as long as the revenue is spent in progressive ways' (which I don't think is present in the blog) is more interesting. However, it misses an important point: the spending does not *just* need to be progressive for this to work. It needs to fulfil the condition that (presumably due to being more efficient) the regressive consumption tax alongside the progressive spending is in total more progressive than the progressive income tax alongside progressive spending. I.e. it needs to be a hell of a lot more efficient to have a leg to stand on.
Then we need to think about whether additional inefficiency is intrinsic to the system itself - where a progressive tax takes money from one person and gives it to a different person (ok, this is oversimplified and it gives it back to the same people as well but you get the point), this system is designed to take money from one person and give it back to the same person.
Finally, there are major concerns to me about raising the requisite level of revenue through a consumption tax. I mentioned the effect this might have on the black market already, but also what you'd be doing is drastically changing the proportion of a poor person's consumption that they have no choice over - it becomes government spending instead of personal spending. Surely this can't be right?
Chris Kav
I think avoidance is still important, Anonymous Economist, but you make a good point about the current levels of incentive affecting the current levels of avoidance behaviour. Others will point to flat taxes as the example for how avoidance reduces when the incentive to avoid income tax is much lower.
Jonty's argument though, is a bit patronising to the poor: if we could just take more of your money and spend it on you, then you'd do better. It's easier than taking it off the rich, you see.
It's an argument for big government, daring other big governmenters to accept to penalise the poor for it, and seems like a bit of a fantasy of those who dislike paying income tax.
1) Income tax was originally, and not just originally, intended as a tax on those with a lot of income. It is through threshholds not tracking earnings that it has come to encompass almost everybody, so that those classed as in poverty are still having their income taken by the government, only to give them back benefits money via a massive bureaucratic machine, demanding humiliating and stressful form-filling.
If we want to be more progressive, then raising the threshhold at which income tax is levied, to the level of a non-poor income, would be the priority. Things would need to be adjusted further up the scale if this were to be neutral for higher-earners, of course, but it makes far more sense to stop taking money from the poor, than to assess them, then take it, then assess them again then give it back. This way, assessment could become something more rare, applied in special cases, rather than applied with blanket effect across such a large slice of the population.
2) Jonty's post misses inequality entirely. Today's poor are richer than yesterday's middle class, and richer than the middle class of some other countries, yet they are poor in this society and they suffer from it.
One big reason for this is inequality, and Jonty's calculus fails to put any price on inequality, even though we know it creates social problems. The supposed greater efficiency bought at the cost of inequality can be canclled out by the problems that inequality brings. It's strange again to accept inequality and the social problems it brings, because one hopes that it will produce the revenue to deal with those social problems.
Government spending is not the answer if the problems might be prevented in the first place. Everyone would prefer that there were fewer people homeless in the first place, rather than more who were provided with government-run hostel-space, or were removed from the streets by greater numbers of Police.
Jonty Olliff-Cooper
Common man,
The amount would be the same in absolute terms but less as a proportion of income.
(To be clear too, the Progressive Conservatism Project is totally independent of the Conservative Party.)
Jonty Olliff-Cooper
Evasion was one of the issues I had in mind.
HMRC estimates that it collects 8% less tax than it is due, a shortfall of £40 billion. Some sources claim the figure is much higher.
The Public and Commercial Services Union estimate that £95 billion is lost through tax evasion and avoidance, http://www.pcs.org.uk/en/revenue_and_customs_group/oracle-magazine/oracle-march-2009.cfm.
In 2007/08, 15% of VAT was lost, 16% of corporation tax and 6% of income tax, due to “errors, evasions and flaws”, http://www.hmrc.gov.uk/stats/measuring-tax-gaps.pdf.
The TUC claims that public purse loses £13 billion a year through tax avoidance by the wealthy and £12 billion a year through tax avoidance by corporations, http://www.tuc.org.uk/touchstone/Missingbillions/1missingbillions.pdf.
Jonty Olliff-Cooper
Economist, Kaveh,
To be clear, I am not proposing this shift: just noting it as an interesting angle on what we take to be a cut and dry issue.
As Kaveh says, the crux is the efficiency of direct versus indirect taxation and that is up for debate, so please do link to evidence on either side.
Indirect tax plus progressive spend would have to outstrip direct tax plus progressive spend for impact. It depends how much more revenue you might be able to raise indirectly than directly, both because it might have a more limited impact on behaviour (not sure on that) and because you may be able to have higher rates because political and public resistance might be lower to taxes that are less high profile than income tax.
I fully understand that indirect taxes are regressive (by the technical definition). However, not all hit the poor heavily if they are not goods consumed by poorer groups. Pitt paid for the seven years war by taxing luxuries like carriages for instance.
Your point about the proportion which you have control over yourself is vital and powerful.
I take your point about the marginal propensity to consume, black markets and the lack of clarity in defining progressive taxation. Not all our readers are anonymous economists, so I was trying to put it in plain terms.
Thank you for taking the time to comment. It is great to have such a sensible and vigorous debate on the blog. Keep posting.
Jonty Olliff-Cooper
Chris, great to have you on the blog.
On your Point 1, I absolutely agree with you. In fact the Progressive Conservatism Project is working on a report on exactly this issue, proposing that we should link the threshold to the JRF’s Minimum Income Standard. Keep watching the website. Should be out in April.
On your Point 2, again, I myself agree. Income inequality is something to tackle regardless. Asset inequality too. The point was more to say, if we had more revenue to redistribute, would that be better than a smaller amount taken more fairly? Depends on the rates of course in part.
I would mention again that this is not something I am proposing. I just found it striking in the meeting and wanted to relay it to the blog.
Thank you again to everyone posting.
Gerty
So to summarise:
"Why don’t we set taxes that punish the poor and hope they won’t notice because it will give them nice schools, whilst rich people also get said nice schools through paying half as much."
Um, no thanks.