Beliefs and arguments concerning the Great Depression reappear every time things get bad enough to raise the question of how to make sure we don’t have another.
The story of 1920-1923 as you’ve presented it is “a recession happened, the government cut spending, and the recession ended.”
But I think if you added a row for 1919, and perhaps 1918, to your table (ideally with CPI and unemployment numbers for each year), the story would look different, to wit: “the Great War ended, the government drastically cut spending, a recession happened, government spending leveled off, and the recession ended.” A story Keynes could have written…
No, it was “level” from 1922-1923, exactly when the recession ended. The big drop in spending was from 1919-1920, indisputably before (and arguably causing) the recession.
And of course Federal spending is planned months before it actually happens, while recessions are declared retroactively, so the 1921 budget was probably set before they even knew there was a recession, not as a response to it.
(a) That sounds as if you were saying that correlation equalled causation. And of course that applies to either side of this disagreement! But in your case, in particular, you seem to be supposing that causation is synchronic, if not instantaneous. Control processes usually act diachronically, and I would expect economic causation to have a time lag of, perhaps, several months if not a year or so.
(b) If we are trying to discuss what effects federal policy decisions had on the economy, retrospectively, then whether those effects were intended is surely irrelevant. Whether the 1921 budget was intended to end a recession, or to cause one, or was set for entirely different motives would not change its effects. (If we could get the economic outcome we wanted by simply having the right intentions, economic policy would be a lot simpler!)
Of course correlation doesn't imply causation... but when you see two things correlated, usually in the same chronological order, it certainly <em>suggests the possibility</em> that the earlier event causes the later one. Which is why I said "indisputably before (and arguably causing)".
And you're right, all these things come with a delay. Changes in government policy take months or quarters to become changes in spending, which take months or quarters to become changes in employment, which take months or quarters to be recognized as a recession or a boom.
I suggested that David add a row for 1919 to his table. I haven't found exactly the source he used, but https://www2.census.gov/library/publications/1935/compendia/statab/57ed/1935-04.pdf p. 164 shows similar numbers, on which basis the missing number for 1919 is at least 18,000,000. (These numbers are in thousands of dollars, so $18B.) The $12B drop from 1919 to 1920 utterly dwarfs the $1.7B drop from 1920 to 1921, or the $1.3B drop from 1921 to 1922, after which spending was fairly flat until 1930.
I haven't found have GNP or employment data for those years, but the CPI seems to have peaked in June 1920 (the last month of FY1920) and bottomed out in August 1922 (early FY23), so we can take those as an approximation of the recession's dates. Those dates are entirely after the $12B spending cut, and mostly after the $1.7B spending cut, which at least <em>suggests</em> that the spending cuts (and the particular <em>kind</em> of cuts — millions of young men leaving the military and rejoining the civilian workforce) caused the recession. And the recession's end, in FY1923, seems to coincide with the end of significant spending cuts, which <em>suggests</em> that flattening the budget helped to end the recession.
I agree that whether a particular economic effect is <em>intended</em> is probably irrelevant. I only mention it because part of David's narrative was that Harding responded to a recession by cutting spending, and I don't believe that's true: 80% of the spending cuts took place <em>before</em> the recession, not to mention before Harding could have been aware of it;. Even the later spending cuts were probably largely military downsizing already planned and committed before the recession was identified.
I wouldn't suggest for a moment that a nation <em>shouldn't</em> cut its military budget after the end of a war, only that it should expect and prepare for the resulting recession, just as whenever a recession ends, we expect inflation.
I think you are right to a degree, that recessions usually follow war. But not because they are caused by the reduced spending that comes from winding down the war effort, but because inevitably with war comes debasement of the currency, and with debasement of the currency comes recession.
you'll note that almost all of the spikes in inflation correspond to periods of war.
Looking at the specific case of the 1921-22 recession, we see a doubling of the money supply between 1914 until 1919. Bust follows inflationary boom, just like what happened in 1929 following the monetary expansion of the 20s.
Oh, no question that war is inflationary: suddenly the government is paying a million young men to do something that produces no consumer goods or services whatsoever, and paying factories to switch their production from saleable goods to military weapons for which there is (we hope!) no market in the usual sense.
But the end of a war also inevitably means lots of former soldiers looking for civilian jobs, all at the same time. Not all of them will get jobs immediately, hence rising unemployment, and the glut of would-be workers causes wages to fall -- both phenomena characteristic of recession, regardless of how much inflation there was during the war.
Actually, https://en.wikipedia.org/wiki/List_of_recessions_in_the_United_States shows a recession for 8 months in 1945, and another of 11 months in 1948-9. The former was indeed expected as a result of the end of wartime spending, and both seem to have been less severe than expected.
Incidentally, the same source lists a brief but severe end-of-war recession in 1918-9, a longer and more-severe one in 1920-1, also attributed to the end of wartime spending, and mild recessions in 1923-4 and 1926-7.
It is also interesting to me, to see what industries contract in a recession and don't recover compared to those business that expand after a recession. The macro impact of creative destruction or destruction and creative recovery? Causation or correlation - I am not sure, but with a big fat research grants we could work on the relationship. Good Post - Thank You
Interesting. I decided not to study econ when my first course used an early edition of Samuelson's book, and our prof told us it would show us how we understood econ well enough that there would be no more depressions. I grew up on a small farm and couldn't swallow that.
The fact that trained economists don’t agree on causes of depression, nor optimal policies, is a strong argument against each of the 3 options you and Huemer argue:
1) Critical Thinking, do some research, think for yourself. — Seems highly unlikely to get to any conclusion at least one or more experts don’t have, so that’s not so much different than
2) follow the experts - tho the way Huemer chooses between competing experts isn’t clear, but presumably it’s at least some (1) Critical Thinking.
Not clearly superior at the personal level with (3) ignore/ NOTA.
Which itself is likely less good than my suggested (4) whatever gives you the most of what you want.
Tho it takes usually some definite benefit to be better than the simple (5) what side are you on, do and support as your tribe does.
I’m more a 1 smaller govt in theory, but it’s better for Rep tribe to win on other dimensions, so I accept sub-optimal huge deficits and winning over better smaller govt and losing.
You can do enough critical thinking to conclude that there is no consensus of the experts, even if the President claims there is. You then either choose your preferred guess on your very inaccurate analysis or accept not knowing.
In my comments on Huemer's post I pointed out that the individually optimal policy is almost always to support the position popular with the people who matter to you and, I think, cited Dan Kahan's evidence that that is what people do.
On one hand, there are subjects where I accept that I don't know, and am not competent to decide which of the experts is in the right, such as interpretations of quantum mechanics. On the other, on many philosophical and political questions, I have spent my life associating mostly with people who don't share my views and often actively reject them. This is at least partly because a lot of my social life involves tabletop roleplaying games, a community membership in which is not based on specific views on those subjects. I suppose membership in the Society for Creative Anachronism and similar organizations may be a similar case; do you find that people there hold political and philosophical views similar to yours?
In the very early years I think there was a meme that West Kingdom knights were left handed libertarian computer programmers, although I may be misremembering the details. Currently, judging both by official policies and Discord conversations, progressives are more common than libertarians.
It wasn't just the recession of 1920-1921. Basically every recession in American economic history prior to the Great Depression was over quite quickly. This was because the government took a non interventionist approach to recessions. Just let them run their course. The result was things got better after a year or two.
Actually, the recession of the very early 1920s was interesting for another reason. This actually lay the groundwork for the interventionist approach that Hoover took in the 1929. Hoover was in Harding's cabinet, and he advocated tirelessly for interventionist measures during 1921. He had pretty much just worn Harding down and was about to be able to implement his program when the recession cleared up and there was no long any need.
For example, he organized a conference on unemployment, bringing together all the great minds of the nation asking how to fix the unemployment problem. Harding gave a spirited defense of laissez-faire at the conference, you can read here:
“One of the reasons I have never become seriously involved in macroeconomics is that it lacks a theoretical structure as solid or as well supported as price theory, popularly but misleadingly labelled "Microeconomics."
Do you expand on this criticism of macro vs microeconomics in any of your articles or writings? I am interested. I was never able to take macro seriously at university.
There is no such thing as a failure of aggregate demand. The problem with the Great Depression is that prices were not allowed to fall. If you keep prices artificially high, then sure, things won't sell. That's not a failure of aggregate demand, that's a failure of keeping prices artificially high.
The reason why the Great Depression lasted so long was because in all the panics and recessions prior to this the government had taken a laissez-faire approach and let things course correct. Hoover, however, embraced "modern scientific economic methods", intervened heavily in the economy, engaged in widespread public works, kept wages high, and the result was to prolong the depression substantially. Of course Smoot-Halley didn't help either.
It looks to me like military Keynesian spending tends to pick winners while normal Keynesians prop up losers. Macho bombast about our military's awesome tacticool goop and sob stories about helping the afflicted may well be skewing my impression.
I believe federal spending went from:
1959 $70.953 $83.102 -$12.149
1960 $81.851 $81.341 $0.510
1961 $82.279 $86.046 -$3.766
1962 $87.405 $93.286 -$5.881
1963 $92.385 $96.352 -$3.966
1964 $96.248 $102.794 -$6.546
1965 $100.094 $101.699 -$1.605
1966 $111.749 $114.817 -$3.068
1967 $124.420 $137.040 -$12.620
1968 $128.056 $155.798 -$27.742
1969 $157.928 $158.436 -$0.507
-and US economic growth stopped in 1970. Requiring every large business to take D patronage hires probably did not help.
Woodrow Wilson had his stroke and no one wanted to take over or draw attention to it, so nothing happened. The Fed did try to stop the natural deflation, but not very well, and that's probably part of why the depression ended so soon.
The story of 1920-1923 as you’ve presented it is “a recession happened, the government cut spending, and the recession ended.”
But I think if you added a row for 1919, and perhaps 1918, to your table (ideally with CPI and unemployment numbers for each year), the story would look different, to wit: “the Great War ended, the government drastically cut spending, a recession happened, government spending leveled off, and the recession ended.” A story Keynes could have written…
It seems odd to me that you call a move from 6.1 million in 1920 to 3.2 million in 1922 "levelling off."
No, it was “level” from 1922-1923, exactly when the recession ended. The big drop in spending was from 1919-1920, indisputably before (and arguably causing) the recession.
And of course Federal spending is planned months before it actually happens, while recessions are declared retroactively, so the 1921 budget was probably set before they even knew there was a recession, not as a response to it.
(a) That sounds as if you were saying that correlation equalled causation. And of course that applies to either side of this disagreement! But in your case, in particular, you seem to be supposing that causation is synchronic, if not instantaneous. Control processes usually act diachronically, and I would expect economic causation to have a time lag of, perhaps, several months if not a year or so.
(b) If we are trying to discuss what effects federal policy decisions had on the economy, retrospectively, then whether those effects were intended is surely irrelevant. Whether the 1921 budget was intended to end a recession, or to cause one, or was set for entirely different motives would not change its effects. (If we could get the economic outcome we wanted by simply having the right intentions, economic policy would be a lot simpler!)
Of course correlation doesn't imply causation... but when you see two things correlated, usually in the same chronological order, it certainly <em>suggests the possibility</em> that the earlier event causes the later one. Which is why I said "indisputably before (and arguably causing)".
And you're right, all these things come with a delay. Changes in government policy take months or quarters to become changes in spending, which take months or quarters to become changes in employment, which take months or quarters to be recognized as a recession or a boom.
I suggested that David add a row for 1919 to his table. I haven't found exactly the source he used, but https://www2.census.gov/library/publications/1935/compendia/statab/57ed/1935-04.pdf p. 164 shows similar numbers, on which basis the missing number for 1919 is at least 18,000,000. (These numbers are in thousands of dollars, so $18B.) The $12B drop from 1919 to 1920 utterly dwarfs the $1.7B drop from 1920 to 1921, or the $1.3B drop from 1921 to 1922, after which spending was fairly flat until 1930.
I haven't found have GNP or employment data for those years, but the CPI seems to have peaked in June 1920 (the last month of FY1920) and bottomed out in August 1922 (early FY23), so we can take those as an approximation of the recession's dates. Those dates are entirely after the $12B spending cut, and mostly after the $1.7B spending cut, which at least <em>suggests</em> that the spending cuts (and the particular <em>kind</em> of cuts — millions of young men leaving the military and rejoining the civilian workforce) caused the recession. And the recession's end, in FY1923, seems to coincide with the end of significant spending cuts, which <em>suggests</em> that flattening the budget helped to end the recession.
I agree that whether a particular economic effect is <em>intended</em> is probably irrelevant. I only mention it because part of David's narrative was that Harding responded to a recession by cutting spending, and I don't believe that's true: 80% of the spending cuts took place <em>before</em> the recession, not to mention before Harding could have been aware of it;. Even the later spending cuts were probably largely military downsizing already planned and committed before the recession was identified.
I wouldn't suggest for a moment that a nation <em>shouldn't</em> cut its military budget after the end of a war, only that it should expect and prepare for the resulting recession, just as whenever a recession ends, we expect inflation.
I think you are right to a degree, that recessions usually follow war. But not because they are caused by the reduced spending that comes from winding down the war effort, but because inevitably with war comes debasement of the currency, and with debasement of the currency comes recession.
https://www.wsj.com/articles/a-brief-history-of-u-s-inflation-since-1775-1450115182
you'll note that almost all of the spikes in inflation correspond to periods of war.
Looking at the specific case of the 1921-22 recession, we see a doubling of the money supply between 1914 until 1919. Bust follows inflationary boom, just like what happened in 1929 following the monetary expansion of the 20s.
Oh, no question that war is inflationary: suddenly the government is paying a million young men to do something that produces no consumer goods or services whatsoever, and paying factories to switch their production from saleable goods to military weapons for which there is (we hope!) no market in the usual sense.
But the end of a war also inevitably means lots of former soldiers looking for civilian jobs, all at the same time. Not all of them will get jobs immediately, hence rising unemployment, and the glut of would-be workers causes wages to fall -- both phenomena characteristic of recession, regardless of how much inflation there was during the war.
I think people expected a recession after WWII on the grounds that we cut military spending, and it never happened.
Actually, https://en.wikipedia.org/wiki/List_of_recessions_in_the_United_States shows a recession for 8 months in 1945, and another of 11 months in 1948-9. The former was indeed expected as a result of the end of wartime spending, and both seem to have been less severe than expected.
Incidentally, the same source lists a brief but severe end-of-war recession in 1918-9, a longer and more-severe one in 1920-1, also attributed to the end of wartime spending, and mild recessions in 1923-4 and 1926-7.
Of course, “regime uncertainty “ is Robert Higgs’ story for the Great Depression. You should cite him if you knew that or read him if you didn’t
I should probably read him. What do you recommend?
https://www.independent.org/publications/tir/article.asp?id=430
I have now added a footnote linking to his piece. Thank you.
It is also interesting to me, to see what industries contract in a recession and don't recover compared to those business that expand after a recession. The macro impact of creative destruction or destruction and creative recovery? Causation or correlation - I am not sure, but with a big fat research grants we could work on the relationship. Good Post - Thank You
Interesting. I decided not to study econ when my first course used an early edition of Samuelson's book, and our prof told us it would show us how we understood econ well enough that there would be no more depressions. I grew up on a small farm and couldn't swallow that.
The fact that trained economists don’t agree on causes of depression, nor optimal policies, is a strong argument against each of the 3 options you and Huemer argue:
1) Critical Thinking, do some research, think for yourself. — Seems highly unlikely to get to any conclusion at least one or more experts don’t have, so that’s not so much different than
2) follow the experts - tho the way Huemer chooses between competing experts isn’t clear, but presumably it’s at least some (1) Critical Thinking.
Not clearly superior at the personal level with (3) ignore/ NOTA.
Which itself is likely less good than my suggested (4) whatever gives you the most of what you want.
Tho it takes usually some definite benefit to be better than the simple (5) what side are you on, do and support as your tribe does.
I’m more a 1 smaller govt in theory, but it’s better for Rep tribe to win on other dimensions, so I accept sub-optimal huge deficits and winning over better smaller govt and losing.
You can do enough critical thinking to conclude that there is no consensus of the experts, even if the President claims there is. You then either choose your preferred guess on your very inaccurate analysis or accept not knowing.
In my comments on Huemer's post I pointed out that the individually optimal policy is almost always to support the position popular with the people who matter to you and, I think, cited Dan Kahan's evidence that that is what people do.
On one hand, there are subjects where I accept that I don't know, and am not competent to decide which of the experts is in the right, such as interpretations of quantum mechanics. On the other, on many philosophical and political questions, I have spent my life associating mostly with people who don't share my views and often actively reject them. This is at least partly because a lot of my social life involves tabletop roleplaying games, a community membership in which is not based on specific views on those subjects. I suppose membership in the Society for Creative Anachronism and similar organizations may be a similar case; do you find that people there hold political and philosophical views similar to yours?
In the very early years I think there was a meme that West Kingdom knights were left handed libertarian computer programmers, although I may be misremembering the details. Currently, judging both by official policies and Discord conversations, progressives are more common than libertarians.
It wasn't just the recession of 1920-1921. Basically every recession in American economic history prior to the Great Depression was over quite quickly. This was because the government took a non interventionist approach to recessions. Just let them run their course. The result was things got better after a year or two.
Actually, the recession of the very early 1920s was interesting for another reason. This actually lay the groundwork for the interventionist approach that Hoover took in the 1929. Hoover was in Harding's cabinet, and he advocated tirelessly for interventionist measures during 1921. He had pretty much just worn Harding down and was about to be able to implement his program when the recession cleared up and there was no long any need.
For example, he organized a conference on unemployment, bringing together all the great minds of the nation asking how to fix the unemployment problem. Harding gave a spirited defense of laissez-faire at the conference, you can read here:
https://www.presidency.ucsb.edu/documents/address-the-opening-the-conference-unemployment-called-secretary-commerce-hoover
“One of the reasons I have never become seriously involved in macroeconomics is that it lacks a theoretical structure as solid or as well supported as price theory, popularly but misleadingly labelled "Microeconomics."
Do you expand on this criticism of macro vs microeconomics in any of your articles or writings? I am interested. I was never able to take macro seriously at university.
I don't think so.
But Milton Freedman did explain the Great Depression: inadequate monetary stimulus to maintain aggregate demand.
And the Great Recession is the same story. Bernanke/Yellen were too cautious (who knows why?) in providing more monetary stimulus.
The points about the inadequacy of fiscal policy are very well taken.
What I am trying to explain is not why they happened but why they were as long as they were. It's only a conjecture — I don't do macro.
There is no such thing as a failure of aggregate demand. The problem with the Great Depression is that prices were not allowed to fall. If you keep prices artificially high, then sure, things won't sell. That's not a failure of aggregate demand, that's a failure of keeping prices artificially high.
The reason why the Great Depression lasted so long was because in all the panics and recessions prior to this the government had taken a laissez-faire approach and let things course correct. Hoover, however, embraced "modern scientific economic methods", intervened heavily in the economy, engaged in widespread public works, kept wages high, and the result was to prolong the depression substantially. Of course Smoot-Halley didn't help either.
It looks to me like military Keynesian spending tends to pick winners while normal Keynesians prop up losers. Macho bombast about our military's awesome tacticool goop and sob stories about helping the afflicted may well be skewing my impression.
I believe federal spending went from:
1959 $70.953 $83.102 -$12.149
1960 $81.851 $81.341 $0.510
1961 $82.279 $86.046 -$3.766
1962 $87.405 $93.286 -$5.881
1963 $92.385 $96.352 -$3.966
1964 $96.248 $102.794 -$6.546
1965 $100.094 $101.699 -$1.605
1966 $111.749 $114.817 -$3.068
1967 $124.420 $137.040 -$12.620
1968 $128.056 $155.798 -$27.742
1969 $157.928 $158.436 -$0.507
-and US economic growth stopped in 1970. Requiring every large business to take D patronage hires probably did not help.
There were two major differences between the 1920 and the 1929 depressions.
Look at https://economics-charts.com/chart-of-consumer-price-index-1800-2005/ and notice inflation occurs during every war and deflation afterwards. Prices and budgets fell so much because WW I had ended.
Woodrow Wilson had his stroke and no one wanted to take over or draw attention to it, so nothing happened. The Fed did try to stop the natural deflation, but not very well, and that's probably part of why the depression ended so soon.