Keynesian fallacies and the Great Depression: or how Australia left Roosevelt eating her dust

Gerard Jackson

This post is a response to a Keynesian reader.

 If the 1931 spending cuts deepened Australia’s depression, as is alleged, then the rate at which unemployment had been rising would have accelerated. In fact, the reverse happened as shown by the chart below. In the year 1928-29 unemployment leapt by 74 per cent and 42 per cent in the following year. For the year ending 30 June 1931 Commonwealth spending peaked at £68,585,546, after which it fell and the Commonwealth began to accumulate surpluses until war broke out. According to Keynesianism this policy should have been an economic disaster. However, as we can clearly see from the chart, not only did the rate at which unemployment had been increasing slow down significantly, rising by only 5.8 per cent, it then began to quickly drop even though the Commonwealth increased its surplus by 277 per cent and cut spending even further.

Manufacturing was extremely important with respect to the demand for labour and accounted for 43 per cent of unemployment. If Nottrampis were right then Commonwealth spending cuts should have caused manufacturing to cut output further and fire workers. Rather than contracting manufacturing employment and output stayed flat for the year 1931-32 after which it rapidly expanded.

Irrespective of what Nottrampis thinks the GDP estimates are telling him the cuts did not deepen the depression in any measurable way. Informed Keynesians know this and that is why they generally avoid the point that Nottrampis thinks he is making. The best that Keynesians and post-Keynesians can up with is the argument that the cuts were deflationary¹. (Spending cuts are no more deflationary than deficits are inflationary).

That national income figures can be an unreliable indicator is a fact that every economist understands, even Keynesians. For example, the 1955 Economic Report of the President states that from 1945 to 1946 the economy contracted by 2 per cent and then rapidly expanded (p. 137). However, Historical Statistics of the United States from Colonial Times to 1970 calculated that from 1945 to 1947 the economy contracted by 12 per cent (p.229). Even better, the National Income and Product Accounts of the United States 1929-1976 reported that from 1944-1947 GNP dived by 17.4 per cent (p. 6).

The last figure exceeds the first two years of the Great Depression while the second figure is almost as bad as the severe 1932 contraction! If this keeps up we will eventually find that the American economy actually disappeared after peace was declared. Joking aside (and that’s a tough thing for me to do when looking at figures like these) the problem is not the deflators that these later statisticians used but the rather obvious fact that they chose not to look at the real economy. Nottrampis has made the same choice. Another point is that it doesn’t pay to treat figures in isolation.

Nottrampis states that “once inflation arrived at last courtesy of the devaluation real wages fell.” Completely wrong. (In fairness to Nottrampis we find Sinclair Davidson and Julie Novac making the same mistake). To begin with the massive drop in imports occurred before the deflation and was brought about by deflation². Recovery didn’t start until some 15 months after the devaluation which was in January 1931. This is just a little too long³. Moreover, the devaluation was not followed by inflation. The next chart shows the relationship between the real factory wage⁴ and the demand for labour.  As the real factory wage fell the demand for labour rose. Now according to the devaluation argument the demand for labour rose because the increase in prices cut real wages. Never happened.

To begin with wholesale prices didn’t start rising until 1934, when the recovery was already underway. But what is particularly interesting is that instead of manufacturing prices rising they fell through 1934 to 1935, flattened out and then started to rise in 1937 while metals and coal prices continued to fall until 1936 after which they too started rising in 1937. (I think I should point out that imports also started to increase in 1934). This is some five years after factory employment started to steeply rise, and continued to do so while the real factory wage, though not the real wage itself, fell. As inflation is usually defined as rising prices it cannot in this instance be responsible for falling unemployment. The main driving force was not devaluation or rising prices but lower costs of production — even though about 75 per cent of capital goods were imported along with a considerable amount of other inputs — combined with greater efficiency in production. In other words, recovery was production-driven, just as it was in America after WWII.

It should be made clear that there is nothing Keynesian about devaluations. Nothing at all. Devaluation is not a deliberate depreciation of the currency in order to export unemployment but a downward correction of an overvalued currency. At one point the premium on the Australian currency against the British pound was 30 per cent. And this was when the pound was also overvalued, indicating that the Australian exchange rate was seriously out of balance with other currencies.

Any classical economist would have advised the Australian government of the time that it was absolutely necessary to adjust the currency downwards in order for the exchange rate to reflect purchasing power parity. Therefore, even if the devaluation was responsible for the recovery it would still not be a Keynesian success story.

On a far less than final note, Nottrampis declared in another comment that “Australia had a weak recovery!!!!!!” I confess to being at a complete loss as to the nature of his logic. Australian unemployment fell from a peak of 30 per cent to about 8.5 per cent in 1938. It peaked at 25 per cent in America and stood at 19 per cent in 1938. In addition, there was never any hidden unemployment in Australian factories (the working week never dropped below 44 hours) while in Roosevelt’s America widespread underemployment in manufacturing was the order of the day.

Making the comparison even worse is the fact that in 1938 Australian manufacturing employed about 25 per cent more labour than it did in 1928. Then there is the little known fact that America also endured capital consumption while Australia did not⁵. Yet Nottrampis declares that Roosevelt’s phony recovery was superior to Australia’s effort merely on the basis of grossly misleading GDP figures. This is the same technique that would have us believe that there was a Great Depression in 1946 but nobody noticed.

Let me further emphasise again the fact that Keynesians categorically state that a policy of cutting spending and running surpluses during a depression or a recession would create an economic disaster. The above charts show this to be nonsense. Australian governments cut spending significantly and ran surpluses right up to WWII. The result was an economic recovery that left Roosevelt eating dust.

Facts don’t speak for themselves: as a rule they have to be interpreted, those charts just might be the exception that proves the rule.

* * * * *

¹C. B. Schedvin, Australia and the Great Depression, Sydney University Press, 1988. p. 252

²L. F. Giblin, The Growth of a Central Bank, Melbourne at University Press, pp. 72, 80.

³This was noted by Schedvin who argued that manufacturing did not exploit the opportunities that devaluation created. This is an implicit admission something else must have been at work (Ibid. 302-303)

⁴The real factory wage is the ratio of the factory to the value of factory output

⁵Arthur Lewis calculated that from 1929 to 1938 net capital formation plunged by minus 15.2 per cent (W. Arthur Lewis, Economic Survey 1919-1939, Unwin University Books, 1970, p. 205). Benjamin M. Anderson estimated that in 1939 there was more than 50 per cent slack in the economy. (Benjamin M. Anderson, Economics and the Public Welfare: A Financial and Economic History of the United States 1914-1946, LibertyPress, 1979, pp. 479-48).