For the AS shock argument to hold true, we would need to see a period of both high unemployment and inflation directly following the shock. However, on first inspection, the data suggests that over the period, rates of inflation and unemployment are moving in exactly opposite direction. The working hours reform occurred, as pointed out by Bowie, at the beginning of 1919. Dowie comments that roughly 91. 4% of the overall reduction in hours made over the two year period of January 1919 to December 1920, had taken place by July 1919.
9 Broadberry implies a "sharp rise in unemployment during 1920-1"10 to emphasise his point that the supply-side shock was ultimately the cause of the depression that ensued. Glynn and Booth disagree with Broadberry and conclude that "mass unemployment was not apparent until 1921". They argue that a lag of 2 years for unemployment to rise in response to the supply shock is too long, commenting that "the 'supply shock' seems to have had a very long fuse.
"11 The data above would tend to support this argument, although other measures of unemployment around the time indicate that there may have been the beginning of a sharp increase in unemployment in 1920 rather than in late 1921 which is implied by Feinstein's data. Examining the trend from figure 4, I disagree with Broadberry's argument that there was a significant rise in unemployment that began in 1920 as opposed to 1921. However, due to the variety of varying data on the period that is available I cannot confirm with utter certainty that Broadberry's calculations were incorrect.
Having established that Broadberry seems comfortable with a one year time-lag for unemployment to respond to the aggregate supply shock of early 1919, if data were to reflect a rise in unemployment earlier than my data suggests, then the corresponding rises of inflation rate and unemployment rates in 1919 and 1920 respectively may give the supply-side shock argument some credit. Prior to the work done by economic historians such as Dowie, Broadberry and Aldcroft on the causes of the 1920 depression, most research pointed to aggregate demand, monetary and fiscal policies as the cause of both the post-war boom and the slump that followed.
Most writers are unanimous in attributing the post-war boom to "demand-pull" rather than "cost-push" inflation. Dimsdale attributes the increase in aggregate demand to a strong demand to rebuild stocks at home and abroad, whilst Aldcroft estimates that consumer expenditure "rose by no less than 21 per cent between 1918 and 1919". 12 Both Aldcroft and Broadberry refer to some sort of build up of consumer demand pressure during the last few years of WWI.
They argue that as a result of few consumer goods being produced towards the end of the war, with all production focused on the war-effort, consumers were forced into saving, instigating a massive upsurge in consumption once the war had ended. This view is, however, disregarded by Glynn and Booth, who point out the lack of any empirical evidence to back this point up. An expansion in aggregate demand (AD1 to AD2), as is implied occurred in 1919 will, ceteris paribus, lead to an increase in the price level (P1 to P2) with a corresponding rise in the level of output and hence employment (fall in unemployment) (Y1 to Y2).
On studying figure 5, we can see that the data in table 1 backs this almost exactly. Between 1919 and 1920 there is a 1. 4% fall in the rate of unemployment and a 4. 4% rise in the price level. This evidence seems to support the notion that the boom was totally demand driven. The causes of the slump are less clear-cut. A popular argument is that tight monetary policy that was employed to try and curb spiralling inflation not only initiated the slump, but played a major part in sustaining it throughout the 1920'2 and into the 30's.
Howson estimates that by the end of 1919 "boom conditions had brought with them something like full employment", as well as well-documented rises in output, industrial production and money income. This did, however, all come about at a cost of high inflation. Having left the gold standard in 1919, due to the unsustainable cost of pegging the pound against the dollar at an artificially high rate, Dimsdale states that "[the] immediate aims of the Bank of England and the Treasury were to check inflation … and to restore control over the domestic money market"13.
In April 1920, just as the economy had reached its boom peak, the Bank Rate was raised to 7%. Dimsdale goes on to suggest that this rise in interest rate caused "a fall in home demand … followed by a decline in exports. " He blames poor judgement in raising the interest rate, saying the rise came far too late to "check the boom", and were sustained for far too long, which only served to worsen the slump. With plans afoot to rejoin the gold standard in the near future, the government were unwilling to change the money supply to try and control the slump, and stayed committed to a policy of high interest rates to curb inflation.
This restrictive monetary policy helped to improve Britain's exchange rate position, with the pound appreciating 20. 8% against the dollar between 1920 and 1922. However, Britain's G. D. P. deflator and weekly wage rates fell by 24. 9 and 23% respectively, in the same time period. 14 Broadberry attributes the fall in price level in the 1920's,solely on the Government's decision to try and restore the gold standard to pre-war parity15
A negative shock to aggregate demand causes the aggregate demand curve to shift to the left. (AD1 to AD3 in figure 5). This contraction in aggregate demand causes a fall in price level and an increase in unemployment. These criteria more than satisfy my data set in table 1 and so a contraction in aggregate demand does look like a more attractive explanation to the slump. Howson and Dimsdale are in agreement that the boom was supported by mass speculation and unfounded consumer confidence in the British economy.
Alford comments "[the] speculative nature of the boom in the latter half of 1919 began to raise doubts in the minds of even the most optimistic businessmen, and it is therefore possible that business confidence was already beginning to wane before the government decided to reverse its policy stance". 16 Therefore, the speculative nature of the boom, that shifted the aggregate demand curve outwards during the boom, could have meant that aggregate demand was forced to quickly contract (shift left) and fall below pre-boom levels, once the speculative bubble burst.
This would explain the speed at which the peak of a boom became a severe depression. Pigou argues that the volatility of the boom-bust cycle of 1919-1922 may have arisen because of the type of goods that were being consumed directly after WWI. Pigou makes a case that there was an increase in demand for consumer durables that did not need frequent replacement, and a mass of what he calls "once-for-all" consumption.
He argues, in line with Broadberry's views, that consumers had a build up of consumption potential in the post -war years and simply chose to buy big ticket consumer durables once the war ended. This would explain sudden and explosive nature of both the boom and the slump that followed it. This is a view that is generally accepted.
Reasons for the sustained slump What was remarkable about the post-war recession was how long some of its effects, most notably unemployment, lasted. Unemployment remained high throughout the 1920's and the 1930's, not falling below 6.8% until 1939. From Phillips curve analysis on the three time periods 1914-1918, 1919-1923, and 1920-1932 (as shown above), I have derived the non-accelerating inflation rate of unemployment (NAIRU) figures for the time period, with very interesting results. The NAIRU figure is a way of looking at the natural rate of unemployment within an economy and shows the level at which unemployment should be to keep inflation constant. It's value, will therefore occur where the change in the inflation rate is equal to zero.