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.
I don’t entirely disagree with those views. In a recent talk I gave on productivity and taxation I started off by quoting both Paul Krugman
Productivity isn’t everything, but in the long run it is almost everything. A country’s ability to improve its standard of living over time depends almost entirely on its ability to raise its output per worker.
and Adam Smith
Little else is requisite to carry a state to the highest degree of opulence from the lowest barbarism, but peace, easy taxes, and a tolerable administration of justice; all the rest being brought about by the natural course of things.
The issue being, what is the meaning of ‘easy’ taxes and ‘tolerable’ administration of justice? If all Wolf is saying that there is more to productivity and long-term prosperity than lowering taxes then he is entirely correct.
But I fear some might interpret his argument as saying there is no relationship between taxation and long term prosperity and productivity and that, I suspect, is not correct.
He shows this graph to illustrate his point.
The first conclusion is that there is no relation between the share of government revenue and the rate of growth of real output per head (that is, productivity) over the 1989-2011 period. The “regression line” is flat.
He doesn’t actually show us a flat regression line – but I can imagine one that is flat and so I am happy to believe him. But looking at his dataset (n = 18), he has OECDitis – those are OECD economies only and there are many more economies in the world and many more in the datasets where he has drawn his data. So I downloaded all the data from the sources he cites and then using his definitions graphed the data and then estimated a regression line (n = 116).
Eviews suggests that the slope of that line is negative and statistically significantly different from zero (p = 0.0005). [Before anyone gets too excited and claims I should do this or that to the data to derive somewhat different results - please remember, this is not my test or regression. I have replicated Wolf's test and regression using a larger data sample.]
So contrary to what Wolf says, there is a relationship between “the share of government revenue and the rate of growth of real output per head (that is, productivity)” once you include all the economies where data are available.
The important thing here isn’t that Wolf has truncated his sample to derive his result but rather that tax itself is only part of the burden of government – wasteful spending and inefficient regulation are also important and the tax debate shouldn’t obscure those issues.