Adding up Ritchie\’s numbers.

Finally, we actually get a reply to the question.

How does Richard Murphy make his sums work about how much revenue will be raised by increasing the tax rate on the top decile of families by 60%?


Good to note you agree the tax could be raised

Your only query seems to be how much tax is raised

Your way of expressing this is to question our assumption on elasticity

For reasons we have given – and which iI have explained here – we think there will be behavioural changes – which may result in more, not less effort

I think there may be a greater demand for work – for example from the second partner in a high earning household that will – in accordance with all observed social behaviour – including the wish to ‘keep up with the Jones’ or to simply pay the mortgage – which is widely ignored in the blackboard economics you favour developed, rather oddly, by economists who have never worked in a market in their lives

That pressure to earn more to cover fixed obligations will, I (we) think counterbalance any tendency to reduce work

Leaving the outcome we predict

And of course we may be wrong

But that’s our prediction

And I stand by it


The effect of any tax change upon revenue is (amongst other thing but these are the two relevant ones) determined by the following.

1) How much people target their post tax earnings as the determinant of how much they work. If people have a number, \”x\”, that they desire, if you tax them more will they go out and work more to get to that post tax \”x\”?

2) How important marginal tax rates are in setting incentives. Do you offer to do more work if the Government is getting 30% of your marginal income or 75% or your marginal income? At what point do you trade more leisure for that decreasing percentage of extra income that you get to keep?

There is really no firm answer to either of these questions. Different people will react differently, the same people will react differently at different times. It is also true that the short term and long term effects are likely to be different. Someone locked into a current spending pattern (say, a particular mortgage, or some number of kids at private school just as examples) may well find that 1) is the appropriate response. However, the same person over time might reorganise life leading to the second effect becoming more important. Trade down the house and take more leisure for example.

(I\’m not sure if I\’ve got this quite right but I refer to those two as the income and substitution effects. Might be me garbling the jargon though….update….and I have, a bit, but not badly enough to affect the argument.)

The nett effect of any tax change is going to be the sum of these two effects.

Richard is assuming in his calculations that the income effect will predominate. That (vastly) higher tax rates will lead to an increase in work and thus an increase in revenue over and above a straight line calculation of that tax change.

That, it has to be said, is an extraordinarily brave assumption to be making. I certainly wouldn\’t want to have to try and defend it. And, sad to say, nor really does Ritchie. He\’s muttered something about the Treasury being allowed to see the sums but not anyone else.

What I do love though is the arrogance of the suggestion that \”blackboard econmoics\” of the type I favour doesn\’t consider such matters.

Gosh, lookee here!

Supply of labour in the short-run: choice for the individual between income and leisure,
including the backward-sloping supply curve.

It\’s part of the A Level economics syllabus. Well, fancy that, I guess \”blackboard economics\” really does ignore such things!

Update: One more thing here. Ritchie is specifically expecting the second part of a household to take up extra work (ie, married women for the most part). But we know that married women are \”more\” sensitive to tax rates than either the single or men. Thus whatever the effects of raising tax rates are generally on either of the two effects, we know that for married women the subsititution effect is stronger than it is for others.

Which really leads us to the conclusion that the specific effect, more married women going out to work as a result of higher tax rates, that Ritchie needs to make his sums balance just ain\’t gonna be there.

11 thoughts on “Adding up Ritchie\’s numbers.”

  1. He appears to have deleted a comment of mine pointing out that conspicuous consumption and positional goods have been known to blackboard economists since a U of Chicago professor coined the phrase over 100 years ago … and my suggestion he tries reading a book.

    just in case he also deletes my subsequent comment, I’ll post it here:

    N.B. Alex might be a “real economist” – I see he describes himself as once having lectured at Oxford – but I think he’s made an elementary mistake.

    “The Frisch elasticity of labor supply is defined as the percentage change in labor supply resulting from a one percent increase in the expected wage rate, holding the marginal utility of wealth constant. Constant marginal utility of wealth means, for example, that an extra £1 is worth the same to someone with nothing or someone with £10m. You can think for yourself what that assumption might do to an assessment of responses to taxation.”

    When you “hold something constant”, that just means you keep it at whatever it is at the point of calculation. So if you are thinking about how somebody earning £10m will respond to an extra £, you take the marginal utility at £10m when making the calc, and if you are thinking about how somebody earning nothing will respond, you take the marginal utility at zero £. This isn’t my area, so I write with some trepidation, but I’m pretty sure that the concept of Frisch elasticity cannot rest on assuming “an extra £1 is worth the same to someone with nothing or someone with £10m” because that’d be absurd and inconsistent with basic neoclassical theory. In general, calculating X holding Y constant absolutely does not mean that Y is a constant. If I’ve got this right, that’s a pretty big mistake for a real economist to be making.

    I think he also completely misunderstand the role of assumptions in economic modeling. Showing how the Laffer Curve emerges from the assumptions does not mean “the existence of the curve is thus entirely based on the mathematical extension of these assumptions” it merely show that such simple assumptions give rise of a Laffer Curve. Consider the simplest neoclassical growth model, the Solow model. This shows us that with a few simple assumptions (constant returns, diminishing marginal products etc.) that you cannot account for the observed differences in income across countries by appealing to differences in savings rates. It would be a mistake to say this insight can be discounted because you don’t like the simple assumptions used to generate it.

    My initial response is also that he’s also wrong about the bias caused by assuming a closed economy – doesn’t the existence of tax competition and mobility make it more likely that raising taxes will reduce revenues? He seems to be saying the authors are finding Laffer effects because of the closed econ assumption. But I haven’t read the paper so may have muddled his argument.

  2. Tim’s analysis of the effect of increased tax rates tends to suggest that those people who would work more (to increase their pre-tax earnings in order to maintain the same post-tax spending power) are those toward the lower end of the earnings range; they need the money so they have to work more, and they are, of course, the only ones who would fill Murphy’s bowl. Those who don’t need the money (because they’re relatively well off) are hardly affected by the increased rate.

  3. “increasing the tax rate on the top decile of families by 60%?”

    Do you mean “by” 60%, i.e. from 50% to 80%, or “to” 60%, i.e. from 50% to 60%?

    We did this with the 50% top rate tax (that increases effective rate to over 60%, once you include National Insurance etc) and established beyond reasonable doubt that it would at best be revenue neutral.

    I’d guess his other mistake is to assume that employers will be creating a load more jobs for all these trophy wives, which, in a high tax economy seems unlikely to say the least.

    Tim adds: Average tax rate, not marginal. From 34% to 55% is his aim.

  4. RM seems to think we are all ants – disturb a nest and they will up their production to handle the group trauma.

    People and ants are different – could someone tell RM ?

    Oh, and as someone who has been seeking work of any kind at almost any pay, for a year or two now, WHERE are these jobs that yuppie spouses can slide into ? Stupid man.

    Alan Douglas

  5. I’ve said it before, and I’ll say it again: we should partition the country and give each a constitution that places a settled political view. In the North, the country can have all those lefty rights. In the South, the country can have rights that protect us from lefties.

    Then we have political cleansing: Ritchie and Polly can accept the settlement in the South or go to the North. Entrepreneurs in the North can come to the South.

    Then we let the experiment begin. We know the result, of course (North vs. South Korea, East vs West Germany). But for the duration of the experiment we have them making trouble only for their fellow travellers and not the rest of us.

  6. Pingback: Liberal Conspiracy » Does taxing the rich work?

  7. Pingback: Blimey

  8. Pingback: Empirical claims can be tested | Tim Worstall

Leave a Reply

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