… I wish to use it on the VAT argument.     Remember that what Policy Exchange put out on VAT was not something like this:

“Did you know that if income taxes were flat and the world was like the following thought experiment then VAT would not be any different from Income tax?”

Instead, from the FT story, which influences the world’s policy makers, we have:

Policy Exchange …  said modelling and academic research showed the idea established in the 1970s and 1980s that VAT was less damaging than income tax was no longer valid.  “We found that, contrary to what is widely assumed amongst journalists and politicians, increasing VAT would be more damaging to economic growth than increases in the basic rate of income tax,” the report concluded. … Politicians should consider restructuring taxes by raising the basic rate of income tax and cutting VAT, it said.

It is a claim that their paper proves this particular point, and as such needs to be taken seriously.  More seriously, I dare say, than a straightforward thought-experiment that clearly misses out essential features of how the tax system works.   The people behind this report are very bright and honourable; they know that they need to treat the world as it is.   Empirical data on how consumption taxes affect growth may help ….

I don’t like arguments from Authority, as you may be able to tell.  People with 20 years  in academia are capable of reaching very different conclusions. Such authority is no vaccine against having badly thought-out views, but merely a great anaesthetic against the stings of criticism that are their best antidote.

But in this case, I find that referring to what the OECD has found in cross-country regressions on tax rates is rather useful.  It is not just “politicians and journalists”, as the PX press release misleadingly implies, but serious economists too.   And the paper is a recent one, not from the 1970s or 1980s.

Here is the table that justifies their view that:

The results of the analysis suggest that income taxes are generally associated with lower economic growth than taxes on consumption and property … These findings suggest that a revenue-neutral growth-oriented tax reform would be to shift part of the revenue base towards recurrent property and consumption taxes and away from income taxes, especially corporate taxes. There is also evidence of a negative relationship between the progressivity of personal income taxes and growth.

The OECD’s findings are about how things work in practise.  Economics as a profession earns its lowest reputation when it is accused of taking a “that’s all very well in practise, but how does it work in theory?” approach.  If your theory produces results that are blatantly at odds to the findings, and go against a lot of conventional wisdom, perhaps the right thing to do is question the rather-too-neat theory.

On the question of a more theoretical approach, here is my second argument from Authority: Greg Mankiw, arguably one of the most influential economists around because of his widely read textbooks, and ex-advisor in Bush’s first term.  A supply sider – someone who cares for growth more than social justice, I should think:

From a strictly economic standpoint, a VAT is great. It is essentially a flat consumption tax, like the so-called FairTax, but implemented in a way to reduce compliance problems. Because it is collected in stages along the chain of production, rather than all at the retail level, tax evasion is more difficult. If you look at the economic effects, a VAT is similar to the Hall-Rabushka Flat Tax, which many economists love. … My bottom line: If I could replace our current tax system (including the personal income tax, corporate income tax, payroll tax, and estate tax) with a VAT, I would gladly do it.

That is from an anti-tax Republican.

Let me be clear: I am not blindly pro-VAT. I recognise the social injustice, and would and will argue for these to be mitigated should it rise again.   But defending it with oversimplified analogies, a blackbox model and mischaracterisations that go against the evidence has not, in my humble view, advanced the argument.

Now back to the Budget ….


23 thoughts on “While I do hate the Argument from Authority …

  1. “like the so-called FairTax”

    Heh, always good to try and influence people on the sly.

    Dunno if i’ve told you this before, but in my first year undergraduate macro class, our tutor said to us: “now you could buy Mankiw’s book – but then, have a think. He was paid a $1m advance for that book – so what does that tell us?”

    The class of 16 stares back blankly.

    “Well, if he gets a $1m advance, the publishers must think it’s going to sell a lot of copies. But the only way to ensure it sells a lot of copies is to make sure the content is of such low standard that any idiot can understand it. Which means the book is unlikely to be very good”.

    Which was a nice introduction for “how to think like an economist”, i’ve always thought.

  2. Though a bit unfair, in that the whole question is begged by the bad=-high-sales argument.

    i should rechristen the blog “a voice of reason asking people not to make up their mind in advance …>”

    1. “Though a bit unfair, in that the whole question is begged by the bad=-high-sales argument. ”

      Why is it unfair?

      If you are writing a text-book with a million dollar advance, then that text book sure as hell better sell a lot of copies. That means it has to be accessible to a hell of a lot of people – which basically means writing a textbook that people on any macroeconomics course at any university in the (english speaking) world can use.

      As a result, what you’re likely to get is something fairly dumbed down.

      So I don’t see what’s unfair about making the sensible observation that Mankiw’s book is unlikely to be very good – or at least, we should be wary of it – given the production background.

      That’s not – as you seem to be implying – blithely and arrogantly “making one’s mind up in advance”, it’s thinking sensibly about more than just the fact that some big-name economist has name on the front of a book, and considering wider factors.

      After all, resources are scarce: first year under-graduates probably only have the money for 1 text book, and there are many on the market. They should think carefully about which to buy. The above reasoning process seems like a sensible one to deploy ahead of making a purchase.

      1. Does a text book being easy to understand necessarily mean that it is of low standard?

        Could it not be that Mankiw is a skilled writer and has gone to great lengths to make complicated concepts accessible to a wide range of abilities?

      2. “Could it not be that Mankiw is a skilled writer and has gone to great lengths to make complicated concepts accessible to a wide range of abilities?”

        Well, perhaps.

        But is it not more likely that he dumbed it down to shift units?

        Bear in mind that when my tutor is telling us this anecdote, we’re also processing the information “he’s an economics tutor, therefore he probably knows whether it’s a good book or not”.

      3. “he’s an economics tutor, therefore he probably knows whether it’s a good book or not”

        … but couldn’t tell you straight up.
        The audience for a textbook is not the student, but the instructor (or departmnet head, etc), who choses whether it becomes the standard text, whether it is mandatory, or whether it is ignored altogether. It is possible that Mankiw ‘dumbed down’ his text in order to impress that audience – after all, he should know how well-recieved a crummy book would be among academic economists. On the other hand, maybe the publisher was acting rationally in attracting a capable author and the author was acting rationally in accepting a nice cheque in anticipation of writng a good textbook (and establishing his bona fides for his next textbook). About the tutor I am not so sure, although striking a pose may be rational behaviour, especially when the penalties are low

  3. Giles,

    You omit to mention what is called a “consumption tax” in the much of the literature on this subject is, in fact, a tax like the UK’s income tax, in which one pays tax in the period of consumption (because of the extensive savings reliefs on pensions, ISAs, houses, etc.).

    I also point out
    (a) that our thought experiment is quite decisive and well-known, reflecting an argument going back to Alan Auerbach and made in the UK context by Stephen Smith. You misrepresent the situation in your other post, by comparing a VAT that isn’t (one that applies to all goods) with an income tax that is (one with allowances and higher rates). Our thought experiment compared idealised income taxes and VAT. Ours is apples vs apples, whilst yours is apples vs pears.

    (b) We explicitly identify the significance of income tax thresholds, using our thought experiment to demonstrate (to *prove*) that it is relative distortions between income tax and VAT that are important.

    (c) We do *not* depend on our thought experiment for our result that basic rate income tax rises are less damaging to growth than VAT rises. After all, it would prove no such thing! It treats them as the same!

    (d) There is a literature on the growth-damaging effects of VAT, which we quote – e.g. Afonso & Furceri 2008.

    (e) Our central argument from the theory is that the relative advantage of VAT over income tax have fallen dramatically since the 1970s and early 1980s, and that the old assumption that VAT is efficient whilst income tax is fair is now obsolete.

    1. Hi Andrew, thanks for the reply.

      I certainly omitted to mention that fact that a consumption tax is really like an income tax; after all, that would be begging the question that we are disputing. While economists may think it is trivial that one is taken from salaries, the other makes prices of certain goods higher, it seems to produce very different effects. And, since the OECD has found such significant differences between the two tax rates to a 1% degree of significance, their decision to treat the two as different taxes seems to have some merit – If I found through a study that one tax had a -0.98 relationship with growth, and another +0.93, I would not look for reasons to find them the same …

      When you say that the thought experiment is decisive, I can agree only in so far as you are indeed comparing apples with apples. One can make decisive thought experiments if one likes – but I am not sure what this then proves. But that prominent piece in the FT which has provoked so much of this thought in me (and no doubt policymaking world) did not claim to have proven that this works in Narnia, but to be good for the UK, now. In which case I agree one would want to compare pears with pears.

      And my VAT was not entirely pears.

      I accept that my simple spreadsheet could not capture VAT exemptions. I agree with Reform (!) that getting rid of these exemptions would be a good idea.
      This would of course be a rise in VAT. Amazing that I am taking Reform’s side against you 😉

      Incidentally, the most striking result I could find in your thought-provoking paper was the one that found such a large effect for raising VAT and dropping NICs. Or raising Income Tax and dropping NICs. I notice that (p52) income tax is treated very differently from NIC’s, with the implication that NIC’s straightforwardly reduce firm profitability, and IT straightforwardly lands on workers’ welfare. Did you think that was realistic? It strikes me that both function effectively like an income tax, and who has to pay them at the first iteration will ‘come out in the wash’ (to quote Tim) once workers and firms have negotiated fully. They are both, after all, labour wedges.

      I must admit to some surprise that the major results from this paper stem from (a) income taxes and NICs being very different (fig 4.7) and (b) VAT and Income tax being very much the same! That seems to explain a lot of the more dramatic results. Would you really be comfortable advising the next Chancellor along those lines? I am curious: I am not all that familiar with the NI system.

      I concede and apologise for not giving greater attention to para top p47. But I also see note 36, which concedes that their analysis is about average and not marginal tax rates. Tax structure seems to matter a lot.

      If you are comfortable advising people that a rise in income tax and a cut of VAT would be good for work and employment, this little blog and 7 minute spreadsheet will not convince you. All I can say is that I still don’t get the theoretical or empirical justification. Best, Giles

      1. Over the long-term, and at the margin, with reasonably competitive labour markets NICs rises will function very much like income tax rises – in either event real post-tax salaries will fall. The question is how we get from here to the long-term.

        In the case of an income tax the mechanism is fairly simple. We reduce the real post-tax salaries of workers. That may cause employment to fall a bit (less work, more leisure).

        In the case of employer NICs the mechanism is more complex. First we increase the costs of employing people. Then firms fire people. Then unemployment bids down salaries. Then at lower salaries people get hired back. And we end up in a new equilibrium in which real post-tax salaries are down by the same amount and employment is a bit down, also (again, less work, more leisure).

        So the mechanism for NICs proceeds via higher unemployment than the income tax mechanism, unless transition paths are smooth and rapid. But in this scenario, in which inflation is v low and salaries falling anyway, it could be quite hard to pass on higher employment costs as lower wages.

        I don’t claim that VAT and income tax are the same in the short-term – only in the longer term. In the short term, there are important differences of distortion. It is those differences of distortion that really count, here.

      2. Thank you Andrew for your as ever thoughtful reply. I hope this blog has not taken too much time from work at PX… it is appreciated (you would be impressed by how much higher the hits/day have been in the last week or two of this).

        I agree with you in acknowledging the rigidities, and the problems of a deflationary environment is that a refusal to take a lower wage may mean that quantities adjust not prices at first. (interestingly, this observation is often the resort of Keynesians calling for higher inflation to reduce unemployment ….)

        However, it is interesting to note how this recession is starting to work against this trend. Who knows why? Responsible unions? Fewer unions? Scarier headlines? I notice The Times last February already noticed that workers were being offered lower pay or the sack. And they seem to be making the right choice.

        I think at the least we would both lobby for a simplified VAT system. Best, Giles

  4. This is a nice piece outlining the implications of the point Giles is making.


    Look at the Nordic tax systems. How do you actually get both a large State, large amounts of social spending and redistribution, and also reasonable growth rates?

    Not that I’m in favour of a large state etc you understand but imagine you were in fact a social democrat who aspired to such things?

    Well, you tax returns to capital lightly (and in some cases not at all), you tax companies lightly, incomes a bit more and consumption quite heavily. As the Nordics do.

    What do the British social democrats (Polly, Murphy, Compass, etc etc) argue for? Low consumption taxes, and high corporate and capital taxation.

    And people wonder why I take the piss out of them?

  5. I thought arguing from authority was just saying “x says y is true and x is an expert in this field so y is true”. So it’s not really an argument from authority if you reproduce the authority’s argument, or point people in it’s direction, rather than just repeating the authority’s conclusion. So actually you’re not arguing from authority so much as using other people’s arguments (perfectly legit)! How wonderful.

    1. I feel strangely comforted! But in future, for the record, I want to say:

      “The OECD thinks this. They employ zillions of economists, they are not idiots and have no political ax to grind. They have a regression with 690 observations. What do you have, sir?” It would save me a lot of bother!

  6. One other thing, Giles. I point you to paragraph 50, p20 of the OECD paper you link to. It says: “Income taxes: This includes categories 1000, 2000 (social security contributions) and 3000
    (payroll taxes) of the OECD classification of taxes.” Thus, it collects together social security contributions (e.g. NICs) with income tax – which, since other studies (not just ours) suggest that NICs damage growth much more than income taxes means it somewhat misses the point in terms of the issues raised in our paper.

    1. I agree that it certainly fails to distinguish those two types of income tax, although I think as you say over the long term NICS do tend towards IT’s. But they do separate out the consumption taxes, which I think is right.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s