Friday, May 18, 2012

A Tale of Two Depressions: 1930s and 1890s Australia

This is a quick follow up to my post on free banking in Australia.

According to the GDP estimates of Noel G. Butlin, Australia in the early 1890s suffered a depression in the aftermath of the collapse of its huge property and financial asset bubble:
Angus Maddison’s Estimates of Australian Real GDP from Butlin (millions of 1990 international Geary-Khamis dollars)
Year | GDP | Growth Rate
1888 | $14,685
1889 | $15,953 | 8.64%
1890 | $15,402 | -3.45%
1891 | $16,586 | 7.69%
1892 | $14,547 | -12.29%
1893 | $13,748 | -5.49%

1894 | $14,217 | 3.41%
1895 | $13,418 | -5.62%
1896 | $14,437 | 7.59%
1897 | $13,638 | -5.53%
1898 | $15,760 | 15.56%
1899 | $15,760 | 0%
1900 | $16,697 | 5.95%
(Maddison 2006: 452).
The moderate recession began in 1890, there was a brief recovery in 1891, but a full-blown depression from 1892 (that is, a period of real GNP/GDP contraction of 10% or more), which continued into 1893. From 1891 to 1893, GDP fell by a shocking 17.11%. From its height in 1889, it plunged by 13.82% by 1893.

How does this compare with Australia’s experience of the Great Depression in the 1930s?

Let’s see:
Angus Maddison’s Estimates of Australian Real GDP from Butlin (millions of 1990 international Geary-Khamis dollars)
Year | GDP | Growth Rate
1927 | $34,716
1928 | $34,164 | -1.59%
1929 | $33,834 | -0.96%
1930 | $32,181 | -4.88%
1931 | $32,720 | 1.67%
1932 | $31,878 | -2.57%

1933 | $33,696 | 5.70%
1934 | $34,991 | 3.84%
1935 | $36,424 | 4.09%
1936 | $38,160 | 4.76%
1937 | $40,336 | 5.70%
1938 | $40,639 | 0.75%
(Maddison 2006: 452).
It appears that a recession already began in Australia in 1928 (this was bad timing), and then the Australian economy was hit by the effects of global Great Depression in 1929. From 1928 to 1930, the Australian economy contracted by 7.30%. There was a brief recovery in 1931, but a further recession in 1932.

These data can be compared:
(1) from 1891 to 1893, under a free banking system, Australian GDP fell by 17.11%. From 1889, it plunged by 13.82% by 1893 (despite the recovery in 1891).

(2) from 1928 to 1930 the Australian economy contracted by 7.30%. Even if one looks at the overall fall from 1927 GDP to 1932 GDP, that fall was 8.17%. Therefore the title of my post is a bit misleading: for the real output contraction of the 1930s did not technically qualify as a depression, just a very severe recession (see here for formal definitions of the terms “recession” and “depression”).
The conclusion is clear as can be: the Australian debt deflationary depression in the early 1890s under a free banking system was worse than its Great Depression of the 1930s!

So much for the superiority of free banking.


Let me repeat Angus Maddison’s assessment of Bryan Haig’s (2001) revised figures of Australian GDP for 1860–1911, and why Butlin’s are very probably better. Angus Maddison points out the following:
(1) for 1860–1911 Haig has no quantitative measure of 70% of GDP (Maddison 2006: 453);

(2) Haig described the estimating procedure he used in but five pages, but Butlin provided his in 200 pages (Maddison 2006: 453);

(3) Butlin provided data for more states than Haig did: Haig used data from Victoria and New South Wales to fill in gaps for overall Australian estimates (Maddison 2006: 453).

(4) one of Haig’s fundamental objections to Butlin’s estimates was that they conflicted with traditional interpretations of Australian economic history: but this is just an unreasonable a priori objection. As Maddison says in reply to this, “it is up to those who disagree with Butlin to prove him wrong” (Maddison 2006: 451).
All in all, I do not see any reason to think Haig’s estimates are to be preferred.


Butlin, Noel G. 1962. Australian Domestic Product, Investment and Foreign Borrowing 1861–1938/39, Cambridge University Press, Cambridge.

Haig, Bryan. 2001. “New Estimates of Australian GDP: 1861-1948/49,” Australian Economic History Review 41.1 (March): 1-34.

Maddison, Angus. 2006. The World Economy: Volume 1: A Millennial Perspective and Volume 2: Historical Statistics, OECD Publishing, Paris.


  1. LK,
    Some questions on your 2 articles:

    What do you mean by estimate GDP? My background in regard to my questions is my experience working as a controller in banking, manufacturing, and distribution for the last 23 years. An estimate is usually based on a forecast, a future event. Or a reserve such as bad debt expense or staled dated inventory. Why would a past event be an estimate? What information is missing in the GDP data to make it an estimate? Are you stating that these estimates are fact or probability? If probability what is the percent probability of factual data?

    Are you stating that the 1890’s Australian depression could have been avoided if the banking system was regulated? If so why was there a depression in the 1930’s under a banking system that was regulated or maybe was not regulated enough?

    How would you apply the Post-Keynesian theory of the business cycle to the 1890’s Australian depression?

    What cause the speculative inflows of capital in the 1880’s? You mention property and stock market speculation. Why did it happen at this time in history and not earlier.

  2. (1) the Australian government keep some data on output from the 19th century, but it was hardly complete. The concept of GDP did not even exist until the 1940s, so therefore we can only ever have retrospective "estimates" of 19th century GDP, on the basis of contemporary record of prices and output volume, with guesses for areas/sectors and regions where data is lacking. These estimates are not certain.

    (2) The 1890s depression could have been avoided with bank regulation, and monetary and fiscal policy, yes.

    (3) the 1930s depression happened at a time where modern Keynesian fiscal policy was not properly developed.

    (4) Minsky's financial instability hypothesis and Fisher's debt deflation theory are, to my mind, obvious the right theories for the 1890s depression.

    (5) capital inflows and outflows in any particular period and country are caused by many factors. Australia became a "fashionable" location for British capital in the 1880s:

    (i) the returns on gold and other resources (silver, lead, zinc, copper and coal) were attractive, and

    (ii) then the returns on property and financial assets during the speculative bubble itself reinforced this. We have all the usual factors related to speculative frenzies, seen from the Dutch tulip bubble to the tech stocks boom of the 1990s.

    Furthermore, even a lot of local capital got caught up in the speculation, so capital inflows are not the whole story.