Tyler Cowen’s Awful Reasoning

Tyler Cowen, in the interests of making the blogosphere waters ever muddier, has written a deeply misleading and confusing post about what exactly constitutes ‘austerity’:

Let’s say that private gdp is 100 and government spending is 100.  Gdp then suddenly goes up to 200, so government spending as a percentage of gdp falls from 50% to 33.3%.  This is not a contractionary event.  It is fully possible to argue “government spending should go up too, to slot more public goods into the larger output,” but the initial change is expansionary, even though government spending as a percentage of gdp took a steep dive.

OK, here’s the problem: this is an absolutely ridiculous scenario that, for all intents and purposes, cannot happen. GDP and government spending/revenue are inextricably intertwined, so an increase in one will almost certainly affect the other. The only way this wouldn’t be true would be if the sector of the economy that grew were completely taxless, and so separate from the other sectors of the economy that it could barely be considered a part of the economy*.

If GDP grows, we expect government spending as a sector of GDP to grow by roughly the same amount**, in the absence of any changes to policy. This is because taxes are collected as a percentage of income, and as income grows governments must pay their staff more, too (I despair that I had to write that sentence). If the government were to pay its staff comparatively less this would be contractionary.

Let’s take the UK economy as an example. The government hasn’t really adopted any major new functions since the welfare state was established post-WW2:

Anyone familiar with UK economic history will recognise the various fluctuations, but overall, spending has hovered around 40% of GDP, even though the economy has obviously grown a lot over the same period. For government spending to decrease as a % of GDP, there would have had to be tax cuts and spending cuts (e.g. what happened around 1980).

There’s also the point that overall spending doesn’t tell us much about what’s going on in individual departments. As many on the left have been pointing out for quite some time, if you cut some areas too fast and during a downturn, you may create unemployment and so welfare spending will go up. This is what has happened in the UK. However, it doesn’t mean that large spending cuts aren’t taking place.

Government spending as a percentage of GDP doesn’t tell us everything, but it’s a good guide as to whether a country is cutting spending or not, even if you do need to factor in tax increases. In any case, it’s certainly better than real or nominal magnitudes with no context whatsoever.

*Insert snarky comment about the financial sector.

*Actually we’d expect it to increase due to Baumol’s Cost Disease, but I’ll put that to one side.


, ,

  1. #1 by Frances Coppola on May 10, 2012 - 6:51 pm

    UE, I’m glad you’ve attacked De Rugy’s ridiculous chart. It was reproduced on the ASI’s blog as well where I had a go at it for the same reasons as you. Government spending is meaningless when separated from GDP. Just as Government debt is.

    ASI link is here – just in case you feel like contributing as well. Their commentary clearly shows the axe they are trying to grind, which is roughly the same as Cowan’s: “Austerity? What austerity?” http://www.adamsmith.org/blog/international/europes-nightmare-has-just-begun

    • #2 by Unlearningecon on May 10, 2012 - 8:14 pm

      I gave up on commenting over at the ASI – it’s absolutely futile. It looks like a couple have had a go, including yourself, but ultimately it falls on deaf ears.

      I’m honestly not sure whether these people are being intellectually honest. I hate to accuse people of things like that just because we disagree but we’ve had this debate so many times that I’m forced to consider it as a possibility.

  2. #3 by James Kroeger on May 11, 2012 - 1:51 am

    So now T. Cowan is arguing that Austerity failed to restore the economy to prosperity because it ‘really hasn’t been tried, yet.’ The simplifying example he produced to argue his case is, as you have effectively pointed out, completely worthless.

    I really do wish that the ‘anti-Republican’ and ‘anti-Conservative’ economists of the world would stop enabling the Champions of Austerity with their meek acceptance of the financier’s claim that (1) tax cuts provide an economic stimulus, and (2) tax hikes are contractionary.

    Both of these claims are wrong, but they can still be found in nearly every economics textbook used in both Europe and North America.

    I find it amazing that economists, who like to think of themselves as serious ‘scientists’, have never bothered to state for the record what the isolated effect is that a tax cut has on the economy, when all other economic variables that could possibly be involved are held constant.

    In other words, cut taxes, but do nothing else (e.g., do not borrow money to maintain government spending). What is the ‘pure effect’ of a tax cut on NGDP? Answer: it is contractionary.

    It is contractionary because governments cannot spend money that they do not have. If taxes are cut and nothing else is done to sustain government spending, then government spending is going to drop by the amount of the drop in tax revenue.

    Because not all of the tax ‘refund’ will be spent (some of the money will be removed from the economy by savers), the increase in C is going to be less than the drop in G. That, my friends, is a contractionary effect, a drop in NGDP caused by the cutting of taxes.

    Raising taxes, on the other hand, is expansionary if at least some of the tax revenues collected by the government would have been saved. The increase in G is greater than the drop in C. Money that would otherwise have been removed from the economy by savers is spent, instead, by the government.

    The result is a net increase in aggregate spending, in NGDP, the very definition of the term ‘fiscal stimulus.’

    It is true that if the government cuts taxes while maintaining its spending levels with borrowed funds, a net stimulus to the economy is generated, but none of the increase in NGDP effected by this fiscal policy ‘combo’ can be rationally attributed to the tax cut. All of the stimulus comes from the spending of borrowed money, money that had been removed from the economy by savers.

    Tax cuts BY THEMSELVES are always contractionary (or at least, never expansionary). But when the government borrows money to finance a tax cut, the stimulative effect of spending borrowed money is usually enough to overcome the contractionary effect of the tax cut (sometimes only barely: see The Bush Years).

    The alternative to Austerity is for the government to reduce its high levels of indebtedness through increased government spending financed by tax hikes on rich people. Indebtedness levels would be reduced at the same time the economy is receiving a direct fiscal stimulus.

    To repeat, the consumption C of rich people would drop, but it would be more than made up for by the increase in government spending G that would occur.

    The empirical evidence available supports this claim that tax cuts are contractionary and tax hikes are stimulative, especially when the the tax cuts/hikes affect primarily rich people. Paul Krugman put together a nice graph that shows this historical correlation.

    Tyler Cowan is ignorant, but what excuse do ‘left-leaning’ economists have for meekly accepting the clever political efforts of right-wing economists to depict their contractionary agenda (tax cuts) as stimulative by combining it with an initiative that actually is stimulative?

    (…and then they turn around and condemn the one thing that makes a tax cut palatable: the borrowing…)

    I can only hope that one day lefty economists will match the political saavy of Republican/Conservative economists and take away their specious “tax cuts stimulate the economy” claim.

    • #4 by Unlearningecon on May 12, 2012 - 10:19 am

      Some good points here – the left are truly awful at framing the debate. It’s always “well, if we just do this to a reasonable extent” or “look, I’m not advocating THAT much government spending”.

      This never works. As far as I am concerned I’m more ready to endorse 100% marginal tax rates than any more cuts.

    • #5 by john77 on May 25, 2012 - 9:05 pm

      Well, I haven’t seen austerity for nearly sixty years, and I don’t think the current government’s fiscal deficit of £100billion or so can be described as “austerity” by anyone with any care for the English language. Osborne is trying to con the bond market but is NOT running a contractionary fiscal policy, let alone going for austerity.
      When you have tried to live on $1 a day, you may have a more realistic view.
      In the abstract the direct effect of a tax cut on the economy is zero: it is the secondary effects that matter and it has been observed that these are generally positive. It is, of course, possible to select special criteria where Usain Bolt or Seb Coe loses [I once finished a few yards ahead of an ex-international because he started half-a-lap behind the rest of us] but Krugman’s specious analysis of the effect of reducing US taxes only on millionnaires is NOT relevant to the British government’s tax cuts for those on below-median incomes accompanied by increased taxes on high-earners. Only Ed Millionaireband thinks that he and his barrister wife are part of the squeezed middle.

      • #6 by Unlearningecon on May 25, 2012 - 10:47 pm

        Of course, your arguments about the definition of austerity have been countered many times – by me above, by Frances, and also by people such as Krugman.

        Me living on $1 a day has nothing to do with it.

        In the abstract the direct effect of a tax cut on the economy is zero

        No abstract concludes this, and you have provided no justification for your assertion. Read James Kroeger’s comment more carefully.

        it has been observed that these are generally positive.

        Again, no; it’s actually the opposite. You should really qualify your assertions better.

  3. #7 by Woj on May 11, 2012 - 6:42 pm


    I agree that Cowen’s example was abstract and somewhat misleading, however I think there is a broader point in the argument that is still worthwhile. My issue with much of the debate is that the term austerity gets thrown around without any context for what each author means.

    You mention
    “As many on the left have been pointing out for quite some time, if you cut some areas too fast and during a downturn, you may create unemployment and so welfare spending will go up. This is what has happened in the UK. However, it doesn’t mean that large spending cuts aren’t taking place.”

    I agree with this statement, but others on the left have been using austerity to mean all government spending and still others don’t mean spending at all (http://bubblesandbusts.blogspot.com/2012/04/what-is-austerity.html). If important areas are being cut, but spending is not decreasing, then it implies that funding is being shifted in the wrong direction. This is probably true but simply pushing against austerity won’t fix that problem.

    IMO the entire debate has become too abstract to offer helpful policy guidance. I think your post is a shift in the right direction but I still hope for an even bigger shift to focus on where spending/revenue is being cut or raised and where it should be cut or raised.

    • #8 by Unlearningecon on May 12, 2012 - 10:17 am

      I think this ties in with the idea of government as a single entity whose size can be ‘increased’ or ‘decreased’. Maybe we even need to abandon the term ‘austerity’ and have our politicians tell us where cuts will fall on a more case-by-case basis.