The Deadweight Costs of Taxation


Someone finally gets this subject into the papers! Jamie Whyte:

Measuring the deadweight cost of a tax system is difficult. You cannot observe all the valuable things that are not made or done but would have been if not for taxes. But economists are clever, and estimates have been made (I will spare you the methodological details). Most put the deadweight cost of raising £1 of tax revenue at between 20 and 50 pence.

Though systematically ignored by politicians, this is a fact of the greatest importance. It means that just to break even, government spending must deliver a return of 20 per cent (or probably more).

Some of it passes this test. The most obvious examples are those where the Government provides public goods: that is, goods that people benefit from even when someone else buys them.

Rubbish collection provides a good example. If your neighbours pay to have their rubbish collected, then you need not. You can simply stuff your rubbish into their bins overnight. Since everyone can figure this out, no one will pay to have their rubbish collected, and soon we will have a public health crisis.

Without tax-funded local council spending on rubbish collection, it might not happen at all. And the return on this spending surely exceeds 20 per cent. The value people place on the aesthetic and public health effects of rubbish collection is far greater than its cost.

But most government spending is not aimed at avoiding such “free rider” problems. Most merely provides people with what they would otherwise buy for themselves, such as education, healthcare, housing, unemployment insurance and pensions. Given the enormous cost of raising funds by taxation, such government spending is ludicrous.

For example, a tax-funded school with an annual budget of £10 million costs society more than £12 million. So, to avoid imposing a net cost on society, state schools must provide education worth at least 20 per cent more than the educations provided by private schools with the same budgets. But how could they?

More than half of all government spending replaces what would otherwise be private spending. So Mr Cameron is wrong that tax is at its acceptable limit. It is at least double what it should be.

Right on Jamie!

There are of course people who have been trying to point this out with reference to the US agitation for tax funded health care, just as an example.

16 thoughts on “The Deadweight Costs of Taxation”

  1. I see a few of the commentators are just ignoring his argument and repeating the mantra that only the state can provide services for the poor.

  2. Personally I think his choice of rubbish collection as an example of a public good is poor. He also then lists other public goods: e.g. education + health as things which people would otherwise purchase themselves. Not generally considered true for socially optimal levels.

  3. There are of course people who have been trying to point this out with reference to the US agitation for tax funded health care, just as an example.

    You reckon a taxpayer-funded system wouldn’t be 20% more efficient than the current US system? Even leaving aside the NHS [*], examples from most other European countries would seem to go against you there.

    [*] I’d argue the NHS was far more efficient than the US system; others would claim that by only letting people have treatments that are proven work and similar levels of MRSA to the US and mainland Europe it’s an Evil Socialist Tool To Murder The Sick; so let’s leave it out for the time being.

    Tim adds: I’m not arguing that it would or would not be that much more efficient. I am arguing that the standard proposers of such systems conveniently leave out those deadweight costs….as indeed we do of the costs of the NHS itself, which, given that it’s funded by taxation, we shouldn’t.

  4. “The Undercover Economist” by Tim Harford contains a good description of the “deadweight costs” with reference to VAT and the on-going example of cups of coffee.

  5. I’ve never quite understood this argument that all taxation imposes deadweight losses.

    The argument is presumably something like this – when I buy a good from a shop, we assume that I get more utility from it than the money, and the shopkeeper gets more from the money than it. A 10% sales tax at the margin means some transactions that would have benefitted us both, doesn’t happen.

    But surely the same is true of the shopkeeper’s profit margin. If he didn’t demand a profit, then the item would be cheaper. Similarly if he didn’t have to pay rent, the item would be cheaper. And in manufacturing the good, if energy wasn’t needed the item would be cheaper. In all these cases there would be more transactions if those things weren’t necessary.

    Now you can say ‘but those things are necessary, and without them there would be no transaction’. But the same could be said for certain government spending. A State is a necessary requirement for transactions to take place, and without it they would not take place, so I don’t think it’s right to talk about such spending needing a higher rate of return as in this example it imposes no deadweight loss.

    Tim adds: That objection is dealt with. There are deadweight costs to everything, as you describe. But if the retailer’s pursuit of a 10% profit makes him more than 10% more efficient than one not doing so, then there is still the cost but it’s not a deadweight loss. Same with having a shop in a building rather than a barrow.

    If government is more productive at supplying whatever it is than the deadweight costs of the taxation to pay for it, you’re correct, there is no deadweight loss. We can trot out the usual classically liberal things here: military, legal system, perhaps some part of the education system, maybe A&E….but there’s always the deadweight cost, the question is whether it is a loss or not?

  6. I find his example hard to swallow as well. Imagine a town of 1,000 families with a local private school to which every family sends their (only) child. This costs £100 each in fees, so the school costs £100,000 a year to run.

    Then the school is nationalised and now every family pays £100 in income taxes.

    Suddenly, we are told, the country is £20,000 to £50,000 worse off, because the business-sapping impacts of the tax.

    But hang on – has much really changed? The idea is that the employee whose was on £10k will now quit his job as he is onlly receiving £9.9k and £10k his reservation wage. But won’t his reservation wage go down by £100 too, as he doesn’t need to pay it anymore?

    Tim adds: Your example doesn’t work. A significant portion of the population don’t have children (most especially of school age). The tax burden falls partly upon them and thus changes their behaviour. Being taxed to pay other people’s school fees is a little different from working to educate your own children, no?

  7. The moment the govt gets involved you have to pay (simplifying things) for a govt building and the salary of a govt employee. I wonder if instead of having the dole – with the enormous staffing/building costs for the army of people who assess claims – we should simply have a basic free meal for anyone who wants one, and the savings from the dole employees would finance it.

  8. It wasn’t clear to me whether the 20% deadweight cost was the average cost or the marginal cost. I suspect it was the average cost. Each additional pound that the government takes from me is worth more to me so the deadweight cost increases with each pound taken in tax, the marginal cost would therefore be higher. The marginal benefit of each pound spent by the government of course declines. In my opinion the marginal benefit of the last several billion spent by the government is currently negative (we’d be better off if they just took the money and burned it).

    I think in Matthew’s example he’s just described a situation where the marginal cost and the marginal benefit are equal for everyone; it doesn’t prove anything because it implicitly assumes its conclusion. And as Tim points out the marginal benefit of any pound of government spending is different for different people – this is true even for public goods.

    Its not clear how much this implies we should reduce taxes. Suppose we should reduce them until the marginal benefit exceeds the marginal cost (including the deadweight cost) for the median (or average? or perhaps those at the 10%ile?) voter. Its not at all clear how much of a reduction this implies; it all depends on how fast the marginal costs decline and the marginal benefits increase.

    I think the only solution is just to appoint me PM, so I can get on with privatising the NHS.

  9. Pingback: IEA Blog » Blog Archive » Obama’s folly: high-speed rail in America

  10. Pingback: US Incomes Are Rising But It’s All Going on Health Care

  11. Pingback: FCAblog » Tesconomics, or how regulation brought us to where we are now

Leave a Reply

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