My Keynesian critic asserts that the “cut in nominal wages which was one of the reasons we got deflation…” This is nonsense. The deflation was triggered when the London funds started restricting credit in order to build up their reserves. The result was a massive contraction from March 1929 to September 1931 that saw M1 drop by 27.2 per cent and demand deposits by a whopping 33 per cent. This should not even have to be said but the cuts in nominal wages were in response to the deflation and in no way contributed to it. Moreover, it is ludicrous to even suggest that a fall in nominal wages could under any circumstances be deflationary. A deflation is a strictly monetary phenomenon1.
He is right, however, in stating that the appalling increase in the unemployment rate was largely due to the increase in real wages. However, his assertion that as other countries got inflation immediately after they devalued then the devaluation of the Australian pound in January 1931 must also have been followed by inflation is flat out wrong2. A picture, as they say, is worth a thousand words. Although Australia devalued in January 1931 chart 1 shows that retail prices continued their fall from 1929 and did not start rising again until 1934, three years from the devaluation.
Wholesale prices started a staggered decline which was not reversed until the second quarter of 1933, some 30 months after the devaluation while manufacturing prices did not start rising until 1937-38, some seven years or so after the devaluation. Moreover, instead of falling, real wages remained basically flat, as shown by chart 2. We can see that the real wage never fell below its 1928 level. Therefore, the assertion that the devaluation generated an inflation that put people back to work by cutting real wages is patently false3. A number of economists and historians have studied this period. To my knowledge not one of them argued that real wages fell. For example, Professor R. G. Gregory, a Keynesian, wrote:
During the depression, real wages measured in terms of consumer prices were above their historical trend. Real wages…did not vary significantly during the 1930s4.
This fact was confirmed by Commonwealth statisticians who stated that
real production per person engaged implies a high real wage for those in employment, and is consistent with available information concerning rate of effective or real wages, which more than maintained in recent years the high level reached in the years 1927 to 19295. (Emphasis added).
Because price indexes consist of a basket of goods they can be misleading. In general raw material prices for manufacturing fell by some 33 per cent or so from 1928 to 1935. For example, the period 1931 to 1936 saw prices for coal and metals, important manufacturing inputs, fall by 14.2 per cent6. Costs were further reduced by the substantial gains that the steel and iron industry made in productivity while productivity in general increased by approximately 1.5 per cent annually. As the real factory wage fell, while the real wage remained comparatively flat, the demand for factory labour steadily rose7, as one can see from chart 3.
The real factory wage (the ratio of the factory money wage to the money value of factory output) peaked at 130 and then started to fall. At this point the rapid fall in the money value of output slowed and factory employment levelled out. (The vertical black lines mark these turning points) As soon as the value of factory output began to rise the fall in the real factory wage rapidly declined.
The result was an immediate increase in the demand for labour that continued to 1938 when there was a slight reversal after which unemployment continued to fall. (In September 1939 the country was at war). So what Australia had was a major reduction in manufacturing costs that led to a production-driven recovery8, all the more remarkable considering the amount of imported inputs that manufacturing used9, that rapidly reduced the unemployment rate. What all of this means, as an economist would put it, is that the supply curve shifted to the right.
There appears to be a gross misunderstanding of the devaluation. The Australian pound was great overvalued against sterling and other currencies. In other words, Australia’s price level was out of kilter with those of its trading partners. The devaluation merely brought the Australian price level into line with other price levels, though it might have overshot somewhat. Therefore, the devaluation brought Australian industry to the stage where it should have been in any case if the currency had not been overvalued.
Another way of putting it is to say that the exchange rate was adjusted to reflect purchasing power parity. Every classical economist fully understood this fact and knew that a devaluation was not a weapon to reduce costs — particularly real wages — but a process of bringing international prices into alignment. Anyone acquainted with the Bullion Controversy would know this. (It should be obvious that there is a fundamental difference between a genuine devaluation and inflation-driven depreciations specifically designed to drive down the exchange rate).
It is also important to note that Australia’s recovery did not begin until about fifteen months after the devaluation. This is a very long stretch for the devaluation to have worked the magic my critic claims for it10. Furthermore, a mere twelve months after the devaluation the Australian pound began to rise again as sterling strengthened. (This puts paid to any suggestion that the full effects of the devaluation were fully maintained throughout the remainder of the decade). In January 1931 the rate against sterling was set at 130. Given that the British pound was also overvalued when Australia tried to maintain parity it is no wonder that the country underwent a massive exchange rate adjustment.
Now in December 1932 the dollar exchange rate was $2.62 but by 1934 the exchange rate had risen to over $4, a rise of more 50 per cent. Imports also started to increase fairly soon after the devaluation. From 1932-33 to 1936-37 per capita imports rose by more than 55 per cent. These facts are needed to put the devaluation in some perspective. They should also disabuse anyone of the notion that the devaluation meant that imports did not increase again during the depression and that the Australian exchange rate did not rise against other currencies.
There is absolutely nothing inflationary about the devaluation process and it is absurd to suggest otherwise. That there is a fundamental difference between how a genuine devaluation works and how inflation operates is something my critic, like so many others who push this line, completely fails to grasp.
No matter what anyone might think the devaluation does not vindicate Keynesianism. Even if the devaluation had been entirely responsible for the drop in employment and the economic expansion it still would not justify Keynesian policies because there is absolutely nothing Keynesian about devaluations. I am against the devaluation explanation for recovery simply because it does not fit the facts, specially when it comes to real wages.
What is truly striking about Australia’s performance during the Great Depression and the thing that really scares Keynesians is that the government cut spending in the depths of the depression by 15 per cent and then ran surpluses until WWII. The result was a rapid and continuing expansion in manufacturing accompanied by a steady and continual drop in an unemployment rate that at one point reached 30 per cent. According to Keynesians this policy should have resulted in an appalling economic disaster.
To sum up: Devaluation did not trigger an inflation and real wages did not fall. Moreover, the evidence clearly shows that the recovery was production-driven. As someone once said: “Facts are stubborn things”.
There will be more on this subject in my next post in which I tackle Roosevelt’s 1937-1938 depression. I shall pay particular attention to the Keynesian absurdity of blaming the contraction on the drop in federal spending.
* * * * *
1Deflation is a monetary contraction. It is associated with swelling inventories, mounting bankruptcies, negative profit margins, widespread unemployment and falling prices. Given these facts it is truly a marvel that a situation where, thanks to technological improvements and entrepreneurial flair, prices continually fall while prosperity rises and real profit margins are maintained can be considered deflationary, even though the money supply is increasing.
2The worst a devaluation can do with respect to prices is bring about a one-off increase. A continuing rise in prices requires an inflationary monetary policy. To argue that the Australian devaluation was the means to lower costs is a serious error and it is one that Sinclair Davidson and Julie Novak made (Institute of Public Affairs: 5½ big things Kevin Rudd doesn’t understand about the Australian economy).
The effect of an overvalued currency is to artificially lower the prices of imported goods relative to domestically produced goods. This means that thanks to her overvalued pound Australian prices in terms of international prices had been artificially increased. The devaluation merely restored the market balance.
3That inflation can increase the demand for labour by lowering real wages is no Keynesian insight. Henry Thornton discussed this very point 210 years ago. He clearly saw that this was done by allowing the money wage to fall behind the general rise in prices. It was not a policy he approved of. (An Enquiry into the Nature and Effects of the Paper Credit of Great Britain, 1802, Augustus M. Kelley, New York 1965, pp. 118, 189-90).
4Recovery from the Depression: Australia and the World Economy in the 1930s, edited by R. G. Gregory and N. G. Butlin, Cambridge University Press 1988, p. 218.
5Labour Report, 1938. No, 29. March, 1940, p.71.
6Official Year Book of the Commonwealth of Australia, No. 32, 1939, p. 424.
7The real wage is determined by dividing the nominal wage by the price level. But employers hire people not according to this wage but the money wage they must pay relative to the value of a worker’s product. Therefore, if the value of the product rises relative to the wage rate the demand for labour will increase.
This thesis is confirmed when the money factory wage is divided by the money value of factory output. The result is an inverse correlation of 0.978. As expected, when the real factory wage fell relative to the value of the output the demand for factory labour rose thereby confirming the standard economic theory regarding wages and the demand for labour. It should also be noted that by reducing the cost of labour relative to the value of the product an increase in productivity would also strengthen the demand for labour.
8Steve Kates has engaged the post-Keynesian Louis-Philippe Rochon in a debate over Keynesianism and aggregate demand. I cannot see how Kates’ approach will sway anyone who is not already committed. You cannot defeat Keynesians without detailed historical examples to bolster your case. Kates is not doing this. He has also failed to grasp the fact that the gross spending approach that Austrians use destroys Rochon’s consumption-drives-demand argument. His response to Rochon’s pent-demand argument regarding 1945 revealed an appalling ignorance of the Keynesian argument regarding America’s post-WWII economy. Worst of all, before the debate had really started Kates conceded a Key point to Rochon: markets are unstable and that’s why we get these booms, busts and financial crises.
9Australia imported the great majority of its capital goods.
10My critic claimed that “Gerry wants to believe the lag associated with the devaluation was the longest recorded in economic history.” Nonsense. I said nothing about historic records. I merely pointed out the very simple fact that the lapse of time between the devaluation and the beginning of the recovery weakens the devaluation argument. No more and no less than that.