Income tax and redistribution

Tim Worstall has an interesting post in which he argues that redistribution is made harder by having big government. I agree with him that:

  • the poor pay too much in tax
  • the state takes an unnecessarily large share of national income
  • income tax should be used as a tool for redistribution

But:

  • he ignores the role of spending in redistribution; public spending – for example on pensions & benefits – is much more important for reducing inequality than taxes. Overall, taking indirect and direct taxes together, the tax system is not redistributive; and all the redistribution caused by government is the result of the redistributive effect of public spending. About half of the government budget is handed back as grants, not spent by Government on goods and services. So if you care about inequality, on the whole you want more, not less, public spending.
  • he asserts (without offering any evidence) that a progressive income tax system is harder to achive if revenues are a larger portion of national income. Presumably the problem is meant to be that it is hard to tax the rich proportionately more than the poor, which is what a progressive tax system requires, as the tax take increases – but it is not clear that this is in fact a constraint on the extent of redistribution through income tax at the current rates of tax.

Both of these make it very unlikely that the size of the Government is in fact an important determinant of the extent of redistribution. I think Tim and I would agree that income tax would be improved if we could broaden the base by reducing tax exemptions and allowances and reduce tax rates – a policy which Nigel Lawson pursued with some success and which this Government has largely reversed. This tends to be progressive because many of the tax breaks and allowances benefit vocal middle class lobbies more than they do the poor. Many people on the right argue for flat taxes, and for tax to shift from direct to indirect tax. Either of these would make it harder to pursue redistribution through income tax, an objective which Tim seems (unexpectedly, to me at any rate) to embrace.

4 thoughts on “Income tax and redistribution”

  1. On the redistribution front it’s rather more “if you believe in it then you should want smaller government” than that “I believe in it”.

    “he asserts (without offering any evidence) that a progressive income tax system is harder to achive if revenues are a larger portion of national income. Presumably the problem is meant to be that it is hard to tax the rich proportionately more than the poor, which is what a progressive tax system requires, as the tax take increases – but it is not clear that this is in fact a constraint on the extent of redistribution through income tax at the current rates of tax.”

    Not quite. We all agree that high (for an unknown value of high)marginal tax rates seriously damage incentives. So we cannot tax the rich (or the middle classes) too much without damaging the general wealth of the nation. Where that value of high is is a different argument.
    This means that there is a limit to the portion of national economic activity that can be taken by the State for its own spending without unacceptably damaging those incentives and thus the wealth/income/growth as a whole.
    If the State is spending more than that (and given that borrowing is just deferred taxation) then the taxation system must be reaching down to take money from the poor to pay for such spending.
    That someone on minimum wage working 29 hours a week faces a marginal tax rate of 30% seems to me to be clear evidence that the State is indeed spending too much precisely because it is taxing the poor in such a manner.
    One could argue that we should be taxing the rich more to pay for current spending levels but I seriously doubt that much can be raised without damaging those incentives (or, of course, depending where we are on the Laffer Curve, higher rates might mean lower revenues).
    Of course I agree with the Lawson points. I would, and do, argue for a Citizen’s Basic Income, the abolition of the rest of the redistribution system, high personal allowances (to get the tax system, including NI, at least above the (relative) poverty level, if not to median income), some agreement as to what are the optimal revenue raising tax rates as compared with their disincentive effects on those who do actually pay in this new system and then cut (or, indeed, expand, but I think that unlikely) government to fit that budget.

  2. Funnily enough, Tim, I suspect that if we could observe the Laffer curve, and if we set tax rates to the revenue-maximising level, it would be at a level which is bigger than you or I would like. I don’t recall seeing any convincing evidence that the maximum of the curve is anything like as low as average tax rates are now – and I wouldn’t be surprised if the tax-maximising rate was as high as 70% or 80%. That would be far too high a tax rate.

    My view is that a more convincing reason to limit the size of the state is not that tax rates are inefficiently high; but that a lot of what the Government spends is a less good use of money than if we left the money in the hands of citizens. In other words, it is not the inefficiency on the finance side, but the spending side, that most worries me. Of course, these are not mutually exclusive effects, and both are good reasons to limit the size of Government: it is just that I think the binding constraint on the size of Government is likely to be when it is doing too much, rather than when it is taxing too much.

    It follows that, in thinking about the right size of Government, I think it is important to distinguish between Government expenditure on goods and services, and Government transfers from one person to another. Both have real economic costs – they have to be financed through taxation which is economically inefficient because it blunts incentives – but it may be that transfers have lower economic costs than expenditure on goods and services (if, like me, you think that there is a cost associated with Government instead of citizens spending money, aside from the cost of getting the money in the first place.) Furthermore, most of the recipients of transfers are themselves taxpayers, so some of the tax and spending is just churning, which may have a different economic effect from other tax-and-spend.

    All of which is a long-winded way of saying that I don’t think the Laffer curve is the right way to determine the size of the state; and that the optimum size of the state depends quite a bit on how the money is spent, so you can’t neatly divorce the decision about the optimum size of the state from the decision about what the state should do.

    But the main point in my original post was the first one. If you believe in redistribution then you are likely to want more spending on benefits, as these are the main mechanism by which the fiscal system reduces inequality. Indeed, because of the regressive effect of indirect taxes, which outweighs the progressive effects of income tax, in most years the tax system as a whole is slightly regressive. So your original point, that people concerned about inequality should want a smaller overall budget, isn’t right. They should want a bigger budget, raised through income tax and spent on benefits for the poor.

  3. One recent attempt to provide a moral basis for the right level of taxation was in Layard’s Happiness. 60% marginal (direct and indirect) he thinks is about right. It’s rather amused me to see so many praising this, thinking that it provides a pretext for more taxation (Guardian columnists notably) when in fact, if you do the sums, it shows that we are currently overtaxed by about 100 billion a year in the UK.

    You might be right that 80% is the tax maximising rate (although I do hope you mean marginal not average rates there) but I seriously doubt it. I’ve worked in an economy where rates were that high (Russia, early 90s)and tax collection, no doubt for a number of other reasons as well, was virtually zero.

  4. I wonder about the differences in jobless benefits from time period to time period. If there are, e.g., beomer/tore/ltnger jobless benefits now than in the past then the incentive to find another job quickly now is less than what it was previously.

Leave a Reply

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