Stimulus Economics: The Data Says NO!

Okay. I’ll admit it. Having read Sinclair Davidson’s post on the Premier’s plan and the stimulus, I was somewhat skeptical. Heck, you have to bear in mind that I went through High School being taught that the Premier’s Plan failed, and besides, I thought that U.S government spending only went up after FDR took office in 1934. Plus Sinclair’s graphs were a little bit fuzzy. So I decided to research this on my own. Because if Sinclair was right, we really had an iron strong case against stimulus packages.

I mean think about it. Take two economies in an identical situation, and test the impact of different government policies on them. It would be the perfect experiment! We could, for instance, find out whether or not stimulus spending actually works or not! And this is exactly what the case was. After all, both countries were in an almost identical position, both followed deflationary monetary policy, the only real difference was the U.S engaged in stimulatory fiscal policy, Australia did not. So. I set to work.

And what did I find.Firstly, I looked up US Federal Spending, which the Office of Budget Management conveniently posts online. And lo and behold, under President Hoover, federal spending did in fact rapidly increase government spending in real terms by over 70% between 1930 and 1932 (from $2.7b to $4.7b, adjusted for deflation), despite revenue collapsing by over 40% ($3.3b. to $1b9b, adjusting for deflation). In fact, by 1934, when FDR took office, US. Government spending had increased a whopping 250% over 1929 levels. In contrast, the Australian government in August 1930 resolved to cut Federal spending by 20%, and restore a balanced budget. Looking at the actual figures for Commonwealth spending from the ABS Year Books, I found that whilst perhaps Federal Spending did not quite fall as much as they wished, there was a real decrease in Federal Spending. I should note that I used the Bureau of Labor Statistics and the Reserve Bank of Australia to factor in inflation (or, as the case was, deflation – how often do you write that!)

Taking this data, I created this graph:


It is clear that 1931 is the year the policies diverge.

From the beginning of the year, the U.S increases spending by 45%, Australia cuts it by 15%. If the fundamental theory underpinning ‘stimulus’ economics holds, we should see, almost immediately, a change in outcomes. Australia’s GDP should decrease, whilst GDP should increase in the U.S. Similarly, unemployment should go up in Australia, and fall in the U.S.

So. What happened. This time I went to the OECD (behind paywall) to find out US and Australian GDP per capita during the early years of the depression, and plotted them such that 1929 was the base year of this index. And guess what happened?


Australia’s GDP starts to grow. In the U.S it continues plummeting. In fact, almost immediately after Australia resolved to slash the size of government, it’s GDP began to rise.

I can not stress this enough. In the year that US Government spending increased 45%, GDP fell a whopping 17%. I repeat, 17%. In Australia, where government spending fell 15%, GDP increased 5%. In the following year, while the U.S “stimulated” economy fell 2%, Australia’s grew a whopping 7%. While the U.S economy eventually begins to improve, as the business cycle kicks in, it does so at a considerably delayed rate to Australia, where the government did not crowd out the private sector thereby delaying the recovery.

There could be no better example of how the economics behind ‘stimulus’ packages fail.

Then I decided to look at unemployment rates, again using ABS statistics and combining them with the BLS. Again, please note, the critical year here is 1931, the year the two countries diverged in policy.

UnempAgain, wow.

From 1931-1932 unemployment index in Australia stabilizes, while in the US it increases almost 8%. Whilst the US rate begins to stabilize the following year, it is clear from the Australian experience that this would have occurred faster, and more effectively, without the stimulus. Not captured here is the fact that Australia had double the U.S rate of unemployment to begin with, and in 3 years had less than half.

I suppose there is a reason why Henry Morgentha – FDR’s very own Treasury Secretary from 1934-1945 said “we have tried spending money. We are spending more than we have ever spent before and it does not work. . . . I say, after eight years of this administration, we have just as much unemployment as when we started . . . and an enormous debt to boot.”

So. To summerise. In 1930 the U.S and Australian governments, when faced with an identical situation, responded in different ways. The U.S tried a ‘stimulus’, Australia tried cutting government.

The U.S. government policy failed. Abysmally. On every indicator.

The facts don’t lie.

My thanks again to Professor Davidson for opening this up to me!

(By the way though, I should note that I don’t claim to be any good at maths, so if I have made any mistakes (and I don’t think I have), please do let me know – I won’t be too offended!)

25 thoughts on “Stimulus Economics: The Data Says NO!

  1. Unfortunately, the correlations are not as meaningful as they might be because they don’t show effects from the coupling of economies. For instance, if current US stimulation “worked”, you could reasonably expect other economies now supplying the USA to benefit – perhaps more and earlier than the USA. However, I doubt if this sort of thing was very material in those days as between Australia and the USA – but for completeness it should still be established. Although it would still be important to adjust for coupling between Britain and Australia then, Britain wasn’t doing much stimulus stuff either.

  2. This is indeed interesting food for thought. What were the relative expenditures are a percentage of GDP? Was Australia already spending big compared to the USA. I do not know the answers to these questions, but I think it would add an interesting extra dimension.

  3. Yewenyi – good point actually; I have all the data for that so should be rather easy to graph up, but shall have to wait till after the weekend (it’s all on my work computer).

  4. PM

    I don’t think coupling for economies is actually that hugely important as the vast bulk of an economy is and has pretty much been domestic orientated.

    Of all people Krugman when he actually was a good economist did some work on this subject in the early 90’s showing that the domestic economy is the thing that matters for the US.

  5. I don’t think that coupling was significant between the USA and Australia in the ’30s, though coupling between the UK and Australia might have been. The point was that completeness needed looking into that. As for Krugman’s earlier work, I suspect it may no longer be as sound, what with the de-industrialisation of “advanced” economies etc. and the associated globalisation.

  6. I don’t think that coupling was significant between the USA and Australia in the ’30s, though coupling between the UK and Australia might have been.

    And the coupling with China and the US is significant while the coupling of Australia and China and the US is significant to Australia and yet we are experiencing a fairly shallow recession.

    The point was that completeness needed looking into that.

    You really don’t have to, as the effects on the external sector of the GDP would reflect that.

    As for Krugman’s earlier work, I suspect it may no longer be as sound, what with the de-industrialisation of “advanced” economies etc. and the associated globalisation.

    You can of course suspect all you like, however he did demonstrate quite methodically to anti-globalist naysayers that the vast bulk of the US economy was internal and the external sector wasn’t as big or nearly as influential as people originally thought. He was quite convincing and I wish he’d concentrate on more of that sort of thing rather than political trooferism he is today.

    You can though, show he was wrong and I’d be only too happy to read your counter.

    So our domestic economy was probably as enormous then as it is now in proportional terms.

    If you’re not going to counter Krugman’s claim you could look at how the Australian economy has fared through this when all it’s major trading partners (which are every bit as influential now as Britain was in the 30’s) have been body slammed further suggesting that it’s still the domestic economy that counts.

  7. Tim Andrews – Good work. You have been busy with excel.

    In terms of looking at economies with comparable monetary policy I think it would be interesting to compare some Commonwealth nations that remained fixed to the Pound Sterling. Perhaps a comparison of Britian and Australia.

    I’ve decided that what Australia needs is a libertarian think tank devoted to Australian historical analysis such as this sort of thing. However not just in regards to economics.

  8. It’d be great to get a book-length treatment of the Australian case from a libertarian perspective.

    Bob Murphy has done an excellent job documenting the American experience in his book. Basically, Hoover and FDR followed textbook Keynesianism AND monetarism – but with dismal results. Also check out his blog.

  9. Actually Yewenyi I take that back, have Australia’s GDP only in USD, and don’t have the exchange rate yet. Can look for it I suppose.

  10. Some selected quotes below on what John Maynard Keynes had to say in 1932 about the Premiers Plan of 1931 and Australian economic policy direction at the time.

    I am sure the Premiers’ Plan last year saved the economic structure of Australia.

    Thus my counsel would be:

    •Reduce budget deficits to a figure allowed by the experts.
    •Satisfy yourselves that the trade balance is adequate to meet pressing requirements
    •Perhaps depreciate the Australian pound 5 or 10 per cent unless sterling itself is allowed to fall a little.
    •Under cover of this, undertake the necessary down ward readjustment of tariffs, which are crippling efficiency.

    Above all, expand internal bank credit and stimulate capital expenditure as much as courage and prudence allow. The substitution of wages for doles needs more credit, but not necessarily much more currency.

    Of particular interest is his point about budget deficits. One might argue that Keynes was actually calling for tax increases rather than spending restraint I wouldn’t find such a claim overly credible.

    You can read his full 1932 article here:-

  11. Terje / Tim – if you look at Julie and my paper released this week you’ll see a graph where we show GDP per capita for the UK, Canada, Australia and New Zealand for the 1930s.

    Tim – If you can access the OECD (behind the paywall) you’ll be able to access Angus Maddison’s historical statistics. Those stats have GDP per capita is constant money units.

  12. Sinclair – yes that’s what I used. But in terms of looking at government expenditure as a percentage of GDP is the problem because Maddison only lists GDP and population, not government expenditure. And the GDP is listed in dollars. And ABS only has government expenditure in pounds.
    So its not finding out the GDP that’s was the problem, its tracking it as a % of GDP.
    In the US you can see it post 1930 at table 2

  13. just comparing the spending patterns of the 2 governments is a bit silly

    in a depression environment the levels of private debt are even more important, and it seems artificial to just assume that deflation and debt deleveraging would influence each economy in exactly the same way (and to the same magnitude)

  14. James – you are right. However comparisons like this is the closest thing we get to a controlled experiment. And Kevin Rudd is arguing that the evidence from 1931 favours his case. When it would seem that it doesn’t.

  15. No it isn’t James. We are seeing if the US Government policy works well, and if Australia’s Government should follow.

    Private debt and equity are of course in important. Government largesse reduces the supply private sector debt. Investment spending drops off quite predictably.

  16. Important for the reading public to know that FDR’s
    (mostly) spend and control policies never got unemployment
    below 14%, and u/t was 17.2% at the beginning of WW2.

    FDR was an admirer of Mussolini (it was mutual).
    The New Dealers copied a lot of their legislation
    from Musso’s program.

    Brian Buckley

  17. This article is brilliant… parallels to the unknown depression of 1920-21 in the US where the government did nothing, actually slashed it’s budgets and the Fed’s market operations weren’t active.

  18. @Terje:

    “I’ve decided that what Australia needs is a libertarian think tank devoted to Australian historical analysis such as this sort of thing. However not just in regards to economics.”

    I wholeheartedly agree… the ease of historical social, political and economic information is preciously low and thus difficult to make arguments for the case of liberty in Australia.

  19. Pingback: Stimulus vs. No Stimulus Historical Comparative Results « The Freedom Thinker

Comments are closed.