It looks like Steve Kates will never get it right on the Great Depression. Harold L. Cole and Lee E. Ohanian wrote a paper blaming Roosevelt’s economic policies for keeping America in depression. Any genuinely informed and honest person would have to agree with them, at least in principle. Now Kates quoted from an article on the work of these two economists that ended with the following quote from Cole:
The fact that the Depression dragged on for years convinced generations of economists and policy-makers that capitalism could not be trusted to recover from depressions and that significant government intervention was required to achieve good outcomes… Ironically, our work shows that the recovery would have been very rapid had the government not intervened.
Steve Kates took immediate umbrage with this view, asserting that “what it doesn’t do is put the blame on public spending which is where the blame truly belongs.” Now there is a fundamental error in Ohanian and Cole’s work but it has nothing to do with public spending, an issue about which Steve Kates is utterly wrong. We get the same nonsense from Sinclair Davidson and Julie Novak who argue that Australia’s recovery from the Great Depression was due to cuts in public spending plus interest rate reductions and devaluation.
Whether he realises it or not Kates is implying that there is either an inverse correlation between the employment rate, industrial production and government spending or no correlation at all. He is evidently unaware that the statistical evidence for America refutes him. Chart 1 shows that as soon as Roosevelt accelerated spending unemployment began to rapidly fall and industrial production began to steadily expand. A Keynesian would need no further evidence in support of his belief that Roosevelt’s big spending program worked. To him, a correlation of 0.879 between the demand for labour and government spending clinches the matter. The poor showing under Hoover can simply be explained away as proof that he did not spend enough. Looked at from this angle Steve Kates’ argument falls to the ground.
The problem is that Kates provides neither a theory nor facts but mere assertions. Davidson and Novak provide some facts but no theory. To merely point out, as they did, that Australia’s unemployment during the Great Depression began to fall when the government started to cut spending is not proof that one is a function of the other. The same holds for American unemployment rates and increased spending. In the case of Sinclair Davidson and Julie Novak a post-Keynesian could argue that the cuts were in fact deflationary and therefore retarded recovery.
What is completely missing from the Kates-Davidson-Novak argument is any reference to real wage rates and prices. Most Keynesians are no better, prattling on as they do about “demand deficiency”. Economics tells us that in a free market there is a tendency for every factor to be paid the full value (discounted value in the Austrian school of economics) of its marginal product. From this we should conclude that the emergence of persistent widespread unemployment is due to wage rates (total cost of labour to the employer) exceeding the value of the worker’s marginal product.
If this holds then a chart would show an extremely strong inverse correlation between wages rates and the value of the workers’ product. Unfortunately, this is not as easy as it probably sounds. Others have tried to solve the problem by using the price level to measure changes in the real wage during the depression and then compare them with changes in the demand for labour. Perhaps a far better method would be one that ignored the price level by taking the ratio of the money wage rate to the money value of the product. Fortunately, the comprehensive statistics for manufacturing allow us to do just that.
If the theory is correct then this ratio would reveal a strong inverse correlation between money wage rates and the value of manufacturing output, producing what we might call the real factory wage. And that is exactly what Chart 2 does, giving us an extremely strong correlation of 0.8291 for the US. As we can see, the correlation covers the whole period. Chart 3 shows exactly the same thing for Australia where we get an inverse correlation of 0.976, which is virtually perfect2.
These correlations confirm precisely what economic teaching predicts in these circumstances. Steve Kates’ assertions that Roosevelt’s increased spending prolonged the Great Depression are dangerous claptrap. It evidently did not strike Kates, any more than it struck Davidson or Novak, that it does not matter for unemployment how much you devalue, cut spending or interest rates if wage rates are kept above the value of the worker’s product. Unfortunately, now that Kates and Davidson have decided that Roosevelt’s spending binge explains the high level of unemployment that plagued the US economy in the 1930s no other view will be considered by our establishment right. And these are the same people who have got the gall to accuse leftists and Keynesians of being close-minded.
* * * * *
1The amount of unemployment in US manufacturing would have been higher had it not been for reduced working hours. Working hours in Australian manufacturing never fell below 44 hours during the whole of the depression.
2In the 1920s Professor Frederic Benham used a similar approach in his study of wages and unemployment in Queensland. The correlation was so strong that he stated: “It would be hard to find a clearer proof of our thesis [that excessive wage rates cause unemployment]”. The Prosperity of Australia, P. S. King & Son, LTD, Orchard House, Westminster, 1928, p.p. 210-12. For some peculiar reason our right refuses to refer to Benham’s work.