31 May 2012

The Taxation Distraction

Martin Wolf, the consistently excellent FT economic columnist, has an excellent blog post up which explains why a focus on taxation as a focus of economic policy debates is misguided.  he writes:
The focus of US economic policy discussion at present is almost entirely on fiscal deficits and the level of taxes. My view is that these are second or even third order issues. What matters far more is the capacity of the economy to offer satisfactory lives for the citizenry. This depends on far more fundamental forces than deficits and taxes, such as innovation, jobs and incomes. Evidently, I am arguing that taxes and deficits do not determine these outcomes. I am suggesting this because they do not.

So I want to address two widely held, but mistaken, views. The first is that lower taxes are the principal route to better economic performance. The second is that the financial crisis is a crisis of western welfare states.
He demonstrates this argument using the figure shown at the top of this post. Wolf lays out the full argument in his blog post, which deserves a full read.

In the US the debate over the role of government takes many forms (and it seems that blog posts on any policy subject eventually arrive there no matter what the starting point or initial direction of travel). The form of the debate at present is framed in terms of a Hamiltonian vs. Jeffersonian approach to government. Snore. It is old wine in new bottle s(in this case the bottles too are old) .

From a policy perspective, the important questions are not simply on how high are taxes and how much does government spend, but how it is spent and what that spending means for growth in GDP per capita. Moving discussion from the former to the latter is, to put it mildly, a challenge.


  1. There is a lot of academic research on the question of the relationship of various forms of taxation to economic performance. It is a complex topic that can't be resolved by plotting a few points. It doesn't seem to me that this blog post contributes to that.

  2. I'm afraid you've been taken for a ride here, Roger. I wondered about the nearly identical GDP growth rates he shows for Denmark and the US, since I knew that we've grown at a much higher rate. See, for example, here:


    Then I noticed he's re-normalizing per capita. Of course, the US has a far higher population growth rate than Denmark. So our GDP growth per capita is correspondingly reduced. Moreover, since most of our population growth comes from immigration and from high birth rates among the poor, the average annual increase in wealth per person in the US is higher than the per capita GDP would indicate. Australia is closer to the US than it is to the Scandinavian countries, all of which have low population growth rates.

    Ireland is indeed an outlier, and one should be skeptical about its GDP, but not for the reasons Wolf gives!

  3. -2-RWP

    Thanks, but did you read Wolf's post? He says:

    "How does one measure economic performance? The most important measure is incomes per head. Employment and the distribution of income matter, too. But incomes per head are the place to start. In the long run, income per head determines the standard of living. So an obvious question is how far tax levels explain growth of income per head."

    I agree.

  4. With regard to the effect of tax increases in the US, Christina Romer and David Romer have written an interesting paper:

    The Macroeconomic Effects of Tax Changes:
    Estimates Based on a New Measure of Fiscal Shocks

    American Economic Review 100 (June 2010): 763–801


    From the abstract:

    The behavior of output following these more exogenous changes indicates that tax increases are highly contractionary.

  5. But he's not plotting average growth in income per head. He's plotting the change in GDP divided by the population. Not the same if people are entering and leaving each country.

    Let's consider an immigrant who entered the US in year 2000 from Mexico. Let's say that immigrant earned the per capita income in Mexico in 1999, and the per capital income in the US in 2000. That person would contribute 0% to (GDP/population growth) in year 2000, but in fact would have experienced a huge increase in his personal income.

    The fact is that most of the low tax countries have GDPs that have grown at a substantially higher rate than high tax countries. They also have populations that increased at a substantially higher rate. (actually, that's interesting in itself). A substantial part of that increase, in the case of the US, Aus and NZ, is immigration. In the case of the US, there are, since 1989, 20,000,000 immigrants who experienced, on average, a huge growth in their personal income, but likely depressed the per capita GDP on Wolf's graph because they were on average less productive, in economic terms, than Americans. Those people are not only hugely better off; they are also cloaking some of the increase in the personal income of everyone else.

    (Reminds me of the joke about the smart Irishman who moved to England and decreased the average IQ in both countries; I'm allowed to tell that one, being Irish)

    One can make a similar argument about people born to families in the lower income percentiles.

    You can't simply ignore population growth. As Jeremy Bentham put it "It is the greatest good to the greatest number of people which is the measure of right and wrong."

  6. -4-RandomReal

    Thanks ... Wolf's argument has to do with absolute values, not rates of change. Thx

  7. -5-RWP

    Thanks ... but 1 million immigrants in a nation of 310 million influence per capita GDP by 0.3% annually ... can't see that as meaningful, sorry. Thanks!

  8. An annual increase from 1.6% to 1.9% in per capita GDP growth over 22 years is actually quite significant. Compound interest, and all that.

    And that's simply accounting for less than a third of the population increase in the US. One has also to account for the fact the higher rate of natural increase in the US leads to a younger and therefore intrinsically less productive population, weighted also toward the low end of the socioeconomic scale. The extra kids we have had since 1989 are contributing almost nothing to the GDP (in fact, they're probably decreasing it), but all are in the denominator.

    Throw in another 0.9% annual GDP growth for the US, NZ and Australia, and I say you have the makings of a trend.

  9. -8-RWP

    Thanks but actually it is 0.3% of 1.6% or on your example, a difference of 1.6% to 1.605%, not much.

  10. In a complex system with many variables, a plot like the one above is probably close to meaningless. It's easily possible that tax rates (deltas or absolutes) have a profound influence on economic performance and population well being. In other words, each of the points may be one possibility on a downward sloping line. Indeed, the data I've studied (Romer was mentioned above, but there's many others now) has convinced me pretty thoroughly that that's the case. The post and the FT article are a huge and (I believe) misleading oversimplification.

  11. Roger,

    Thanks. The chart above plots growth (rate of change) and not absolute value. If one does look at absolute value (in the unmarked second chart), some of the countries with the lowest tax rates have the highest GDP/capita. While I agree that overall tax burden is basically uncorrelated with GDP/capita growth, it's a difficult job comparing countries. I suspect that these measures are strongly influenced by the structure of the taxes and spending as well as the economic history of the country. I pointed out the Romer paper because it was a study of the effects of taxation on GDP growth purely within the context of the US. They were very thorough in their analysis. It should certainly be taken into account when deciding policy, along with a gazillion other factors. I'll leave that nightmare to the politicians (who generally do whatever they want) and the economists (who will fight the Jefferson/Hamilton battle for the next few millennia).

    Great blog. Armchair economist mode - off.

  12. To call discussion on taxation policy a distraction sounds exaggerate to me.

    But my attention was drawn by his graph trying to convey 20 years economic history as a single average. I plotted the "phase-space" trajectory of these random variables (I've used gov. consumption instead because that's what I've had handy) and it suggests to me that low taxation economies seem more volatile (exceptions are US and Switzerland).

    The figure I produced is here


    The name of the countries lies at the start of the series (1989) and axis units are % (sorry is a bit messy, but that's what I can do in 15min)

  13. No, Roger. If you assume that immigrants have had negligible effect on the GDP, as per the Smith and Edmonston (1997), but added 20,000,000 to the jobs total since 1989, then therefore they have reduced the per capita GDP growth from 1.9% to 1.6%. Do the math.

  14. -13-RWP

    The GDP/capita rate shown in the figure is an annual rate, not a cumulative rate.