With the United States’ economy seemingly “recovering”, comparisons of the years 1929–1932 to those of 2007–2009, and the current period of recovery to the time period spanning 1933–1939, were easy to make. It was only a matter of time before economists began to look for a new comparison—the recession of 1937 to a potential “double-dip” today. The 1937 recession was preceded by a decrease in deficit spending and an increase in the reserve requirements of banks by part of the Federal Reserve. Otherwise known as “Roosevelt’s Recession”, it is a perfect example of how a drop in government spending and tight monetary policy leads to economic disaster. Or so goes the official story.
The actual lessons of Roosevelt’s Recession are much different. The 1937–38 dip was not the product of tight fiscal and monetary policy, but of excessive government regulation and loose monetary policy. The misconception that without government deficit spending and easy credit the market will freeze must be cleared, or else we risk allowing the Government and its financial institutions to further disjoint the economy.
The official story tells us of a recovery which took place upon the ascension of Franklin Delano Roosevelt to the presidency of the United States. He replaced penny-pinching Herbert Hoover, who foolishly trusted in the power of the free market to correct itself and self-regulate. Roosevelt immediately began to pour money into public works projects, social insurance programs and other types of government welfare. While the government accumulated a substantial government debt, thanks to Roosevelt’s efforts, the economy recovered at a respectable rate between the years 1933 and 1936, avoiding perpetual poverty and suffering.
In 1937, Roosevelt decided to balance the budget and drastically cut deficit spending. Nearly simultaneously, the Federal Reserve raised reserve ratio requirements for member banks, leading to a contraction of the monetary base. The results were tragic as the economy slumped back into recession. The economy was not strong enough to support itself, and tight fiscal and monetary policy allowed a fickle market to waver in the face of what was otherwise a stable period of recovery.[1]
Christina Romer, Barack Obama’s current chair of economic advisors, sums up the official history:
[T]he recovery in the four years after Franklin Roosevelt took office in 1933 was incredibly rapid. Annual real GDP growth averaged over 9%. Unemployment fell from 25% to 14%. The Second World War aside, the United States has never experienced such sustained, rapid growth.
However, that growth was halted by a second severe downturn in 1937–38, when unemployment surged again to 19% … The fundamental cause of this second recession was an unfortunate, and largely inadvertent, switch to contractionary fiscal and monetary policy. [Spending cuts and tax hikes] reduced the deficit by roughly 2.5% of GDP, exerting significant contractionary pressure. [2]
Oftentimes, the official history turns out to be wrong with the presentation of previously unconsidered facts. This history of the Great Depression is not an exception. The opinions that Herbert Hoover was a laissez-faire president and Roosevelt’s New Deal paved the road to recovery have been refuted elsewhere.[3] Here we shall concern ourselves only with the myths of 1937.
Myth 1: Deficit Spending
That the deficit in 1937 was smaller than that of 1936 is undeniable. In 1937 the deficit stood at $2.193 billion, as compared to a deficit of $4.3 billion in 1936. However, also noteworthy is that while the deficit was half as much as that of the previous year, total government outlays decreased from $8.228 to only $7.58 billion.[4] Interestingly, during 1935—a year considered one of recovery—total government outlays measured $6.4 billion, less than during both 1936 and 1937. In fact, looking back to the years from 1933 to 1935, government spending peaked at $6.5 billion in 1934. It suffices to say that explaining Roosevelt’s Recession by pointing at a decrease in government spending is severely dishonest.
It is not much more useful to look at deficit spending, either. True, deficit spending in 1937 was at its lowest since 1933, but it is worth mentioning that in 1938—the same year as the economy rebounded from the 1937 dip—total government deficit spending amounted to only $89 million! One can conclude that looking at deficit spending is at best misleading and at worst as dishonest as blaming the recession on a decrease in government outlays.
But wait, if government spending did not decrease by much, then how did the government avoid large deficits? Government receipts—money received through taxes—increased from $3.9 to $5.4 billion between 1936 and 1937. In other words, high government spending did not result in a high annual deficit because the government collected a far greater amount of tax money that year than in all previous years of the Great Depression. It is unsurprising to note that in 1938 government receipts increased to $6.75 billion.
While tax hikes played a part in damaging any recovery which could have taken place, it makes little sense to single them out only in reference to the recession of 1937.
Finally, while government spending did decrease between 1936 and 1937, it holds that total expenditure in 1937 was still greater than all years prior to 1936. If contractionary fiscal policy led to a recession in 1937, how did less spending cause recovery only a few years before? This inchoate theory does not hold water.
Myth 2: Tight Credit
Blaming tight monetary policies on cyclical fluctuations has been a favorite ever since Milton Friedman’s and Anna Schwartz’ extensive, albeit heavily flawed, monetary history of the Great Depression.[5] Why such a terrible depression in 1929? Restrictive monetary policy, of course![6] Why the recession of 1937? Restrictive monetary policy! This theory applied to 1937 is no less inaccurate than when applied to 1929.
The 1937 recession came with a contraction in the money supply. The reasons for this contraction are placed upon the shoulders of a restrictive Federal Reserve, which raised the reserve ratio in the previous months.[7] This reason is unsatisfying, because immediately after the reserve ratio was raised the volume of loans made and securities sold continued to rise. It was only after the initial fall in the stock market that bank lending began to tighten.[8] A more plausible explanation behind the contraction of the money supply was a tightening in lending and a decrease in borrowing due to an increase in entrepreneurial uncertainty, given the drop in the stock market’s value and growing disproportion between real wages and productivity. The fall in the supply of money was a result of the 1937 recession, not vice versa.[9]
If anything, a loose monetary policy preceding the recession was more at fault than the Federal Reserve’s ineffective attempt to lower excess reserves by raising the reserve ratio.
Causes of Roosevelt’s Recession of 1937
If it was not tight credit or low government spending, then what caused the 1937 downturn? Three main factors stand out. An inflow of gold from Europe and an artificial increase in the dollar–gold exchange ratio caused a period of inflation. Meanwhile, government union and wage policies maintained high real wages in the face of stagnating productivity. Finally, heavy government regulation made the stock market extremely volatile and susceptible to otherwise minor changes.
1933–36 saw an expansionary monetary policy pushed by the Federal Reserve and the Federal Government.[10] Within that time period, the stock of money increased by 46-percent and the general price level by 31-percent.[11] While the Federal Reserve’s rediscount rate remained at 3½-percent for the majority of that time,[12] the greatest monetary inflation came as a result of the inflow of gold from Europe. Regime uncertainty in Europe, largely as a result of the rise of Adolph Hitler in Germany, caused an influx of gold into the United States. In 1934, the government increased the price of gold from $20.67 to $35 per ounce. Banks holding this increased stock of gold were keen on exchanging it for dollars, given the artificially high exchange rate, leading to a substantial increase in the money base.[13]
We know from the Austrian business cycle theory that increases in the supply of money will lead to “shifts in the structure of production”.[14] This means that there will be a shift towards the production of capital-goods, as they show to be temporarily advantageous while the interest rate—or cost to borrow capital—is low.[15] This occurs because lower interest rates imply the share of profits made from investment of capital-goods will increase, given that the cost of borrowing the necessary capital is lower than otherwise.[16] The result is widespread malinvestment, as the decrease in the rate of interest was not preceded by a necessary increase in the volume of voluntary savings. Such was the result of the artificial increase in the price of gold in 1934. It comes as little surprise that between 1935 and 1936 there was a sudden illusionary boom in productivity.[17]
Although real wages decreased at first, by 1937 real wages rose by 11.6-percent. It is no mere coincidence that around that time the Supreme Court upheld the National Labor Relations Act of 1935, the Social Security Act and the Wagner Act. The result of these decisions was an increase in the power of unions to coerce firms to raise wages and benefits.[18] Fringe benefits, or supplements to standard wages, rose from 1.4-percent in 1935 to 4.2-percent in 1937. Accounting for the majority of the rise in cost of supplements was the required employers’ contributions toward social insurance, which rose from 25 to 71 percent of the total cost of the supplements, by 1938.[19]
It is also important to consider Friedrich Hayek’s “Ricardo Effect” theory. This is the tendency for entrepreneurs to replace labor with capital-goods while the productive structure lengthens, due to increases in real wages.[20] The period between 1935 and 1936, as previously explained, saw an economic boom, a lengthening of the productive structure. But, union activity largely disallowed entrepreneurs to replace labor with capital-goods, while real wages were rising astronomically; by 1937 real wages surpassed increases in productivity.[21]
All that was necessary was a catalyst to bring about a slowing in the pace of credit expansion. This was provided by the stock market. Heavy regulation in the years leading up to 1937, including heavy taxes and legal impediments on inside trading, reduced incentives to invest in the market. The result was a stock market in which a large proportion of shares were held by a relatively small pool of investors. This naturally “thinned” the markets, meaning that minor changes relating to buying and selling could reflect dramatically on the prices of individual stocks.[22] Regime uncertainty caused by several of the Supreme Court’s decisions and increased tax rates consequently led to volatile fluctuations in the stock market, as investors moved to sell their shares.
The sudden drop in value of aggregate stock indexes led to more widespread uncertainty, causing a decrease in the volume of lending. This catalyzed a contraction in the credit markets, and the widespread malinvestments which occurred in 1935 and 1936 began to reveal themselves. The 1937–38 period, known as “Roosevelt’s Recession”, was therefore a necessary readjustment period after the boom of the period 1935–36. More roundabout methods of production, or more capital-intensive entrepreneurial activities, were found to be less profitable than beforehand believed. Moreover, marginal profitability was undercut by the high cost of wages. The predictable result was an incredible drop in productivity and a rise in unemployment.
It is evident that the recession of 1937 was not a product of low government deficit spending or contractionary fiscal policy on part of the Federal Reserve. It was, instead, a product of expansionary monetary policy and heavy government regulation. These are important lessons to learn, as we face a new period of recession and slow recovery. We should not be fooled into believing that without high government spending and loose credit organization the economy will crumble. The actual situation is far different. Indeed, high government spending and easy money only promise to increase malinvestments and forestall necessary readjustments, promising a decline in prosperity.
[1] The case against the Federal Reserve’s tight monetary policy was defended by Milton Friedman and Anna Schwartz in A Monetary History of the United States, pp. 493–545. Also worth reading: Roose, Kenneth D., Federal Reserve Policy and the Recession of 1937–1938, The Review of Economics and Statistics, Vol. 32, No. 2: May 1950; p. 178.
[2] As quoted in Murphy, Robert, “Christina Romer’s Faulty Depression History”, Mises Daily: 6 July 2009.
[3] Rothbard, Murray, “Reliving the Crash of ‘29”, Mises Daily: 21 December 2009. Also worth reading: Murphy, Robert, The Politically Incorrect Guide to the Depression and the New Deal, Regnery, 2009.
[4] All budgetary information (outlays and receipts) is provided by the Government Printing Office.
[5] Op. Cit.
[6] For a refutation see: Rothbard, Murray, America’s Great Depression, BN Publishing: 2008.
[7] Friedman and Schwartz (1963) and Roose (1950).
[8] Anderson, Benjamin M., Economics and the Public Welfare, LibertyPress, Indianapolis, Indiana: 1979; p. 432–434.
[9] Salerno, Joseph T., Money and Gold in the 1920s and 1930s: An Austrian View, The Freeman: October 1999 (Vol. 49, No. 10).
[10] Ibid.
[11] Friedman and Schwartz (1963), pp. 497–498.
[12] Ibid., p. 512.
[13] Anderson (1979), p. 346.
[14] Hayek, Friedrich A., Prices & Production and Other Works, Ludwig von Mises Institute, Auburn, Alabama: 2008; pp. 78–79.
[15] Ibid., p. 67.
[16] Huerta de Soto, Jesús, Money, Bank Credit and Economic Cycles, Ludwig von Mises Institute, Auburn, Alabama: 2009; p. 349.
[17] Higgs, Robert, Depression, War and Cold War: Challenging the Myths of Conflict and Prosperity, Independent Institute, Oakland, California: 2006; pp. 11–13.
[18] Vedder, Richard K. and Gallaway, Lowell E., Out of Work: Unemployment and Government in Twentieth-Century America, New York University Press, New York: 1993; pp. 130–131.
[19] Ibid., pp. 140–141.
[20] Huerta de Soto (2009), pp. 330–331.
[21] Salerno (1999).
[22] Anderson (1979), pp. 441–442.
On Myth 1, RE: “It suffices to say that explaining Roosevelt’s Recession by pointing at a decrease in government spending is severely dishonest.”
That’s why nobody explains it that way Jonathan! The key is deficits, not spending. If you just finance spending with taxes you’re just trading consumption and investment for government spending. What have you gained? Nothing. And, as we’ve discussed in previous posts, it’s DEFICIT SPENDING specifically that gets the economy traction during a liquidity trap, NOT government spending. So, I hate to break it to you but your entire Myth 1 is a non-sequitor. Also, check your dates on recovery and the deficit – the economy didn’t start to recover until mid-1938, when fiscal year 1938 was almost done. Employment took even longer to recover than production.
I was also extremely curious about this point that you made: “The result is widespread malinvestment, as the decrease in the rate of interest was not preceded by a necessary increase in the volume of voluntary savings.” You seem to imply that investment demand is stable. Why else would you insist that the interest rate has to respond to a change in savings? What reasons do you have for assuming that? You seem to be assuming your own conclusions – i.e., assuming that investment demand is driven by a given interest rate (given either by loose monetary policy or savings behavior, that is).
Daniel,
Within the Keynesian framework, government spending is meant to replace consumer spending and private investment where it doesn’t exist. This is because people aren’t spending their money. Whether the government is spending through deficit, or it’s taxing and spending, it is completely irrelevant. You are muddling the concept, and I don’t think you are thinking this through as clearly as you could.
If what you are saying is true, then it would also be true that in the event of normal economic growth and high private investment, government deficit spending would still be beneficial. This is clearly untrue, which is why during periods of economic growth governments generally scale (or try to, at least) back spending and allow the private sector to invest and consume with as little intervention as possible (through a decrease in taxation, for example).
In times of economic distress, the Keynesian argument lies solely in spending (even deficit spending will have to be paid through taxation, at some point). Looking at deficit spending is superfluous, and to claim that overall spending is irrelevant while only looking at deficit is important is being disingenuous (for reasons previously explained). If private investment is low, then it doesn’t matter if the income is being taxed, because it’s not being spent anyways. At least, that should be the Keynesian argument.
As for your second point, it’s assumed that the reader has some understanding of the Austrian Business Cycle Theory. If you need me to explain why savings and investments should correlate, then do some research on the Austrian Business Cycle Theory. This article was not meant for that.
Also, on the note of when the recovery took place, and when government expenditures began to increase, it’s interesting to see how today Keynesians will claim that there is a time lag between when money is spent and when the results are seen. Are you dropping that claim? Because, if that claim is true, then the first spending relevant is that which took place in the fiscal year of 1938.
RE: “This is because people aren’t spending their money. Whether the government is spending through deficit, or it’s taxing and spending, it is completely irrelevant.”
You could not be more inaccurate. Name me a Keynesian that would advocate a tax-financed government spending.
RE: “If what you are saying is true, then it would also be true that in the event of normal economic growth and high private investment, government deficit spending would still be beneficial.”
I’ll leave aside “functional finance” arguments that justify a perpetual deficit (which I agree with), and assume you mean large deficits in normal times should be as beneficial as large deficits in bad times. This is wrong as well. In normal times, government would compete with firms and households for financing. If that is the case, then the allocative efficiency of the market is always going to surpass the allocative efficiency of the government (except of course in certain public goods), and large deficits would hamper growth. Deficit spending only makes sense when investment demand is depressed.
RE: “As for your second point, it’s assumed that the reader has some understanding of the Austrian Business Cycle Theory. If you need me to explain why savings and investments should correlate, then do some research on the Austrian Business Cycle Theory. This article was not meant for that.”
I’m not sure what point of mine you are refering to when you say “savings and investments should correlate”, but I do understand that this point of yours is implicit in ABCT. My point is that if your conclusions are dependent on your assumptions (as they are in this case), you need to be more up front about the fact that that is the assumption you are going in with. The analagous Keynesian assumption (which I only partly buy into) is that investment demand is insensitive to the interest rate. It’s the reverse of your argument, in a lot of ways. When the conclusions of Keynesians is derived largely from this assumption, they need to make explicit that it is the assumption that drives the conclusions, rather than the logic. I’m just asking something comparable of you.
RE: “Also, on the note of when the recovery took place, and when government expenditures began to increase, it’s interesting to see how today Keynesians will claim that there is a time lag between when money is spent and when the results are seen. Are you dropping that claim? Because, if that claim is true, then the first spending relevant is that which took place in the fiscal year of 1938.”
I’m not sure what you mean. Some of the impact is going to be immediate, but yes there will be a lag in the full impact because of the multiplier. So? The deficit started shrinking before FY38, the economy started shrinking sometime in ‘37, and the impact on employment lingered long after production had resumed. I’m not one that thinks that the recession within the recession was entirely due to fiscal tightening. A lot was going on then – including the supply shocks that you raise that are related to unionization. But I don’t see anything inconsistent about my understanding of the lagged effect of deficit spending and the behavior of the economy.
Daniel,
Don’t be naïve. How do you propose government should finance spending? How do you think that debt is paid for over the long-run? Do you think that the accumulation of debt can be done so for extended periods of time? Even Keynesian economists such as Paul Krugman agree that debts have to be paid for over the long-run (of course, what Paul Krugman doesn’t account for is that the deficit spending has to be scaled down for this to happen). Do you think that the government can pay for programs without taxation? Even if the government pays for programs through inflation, this just manifests itself as a tax on all people, and in this case it’s worse on the poor.
Furthermore, please stop confusing a “perpetual deficit” with a “perpetual deficit growing at an accelerating pace”. My suggestion would be for you to map out your beliefs and your opinions, and make sure that your final thoughts actually take into consideration the topic being discussed. That way we can avoid arguing about things which are not even completely relevant to the debate.
In any case, thank you for proving my point. You write:
Thank you for agreeing with me. I will quote what you said in your first post:
I hope you realize where your contradiction lies. By all accounts, working within the framework of Keynesian theory, if the money is not being spent, then if it is taxed and invested by the government, then it is not competing with private expenditure. So, within the framework that you yourself operate in (but, ironically, have an inferior understanding of), government expenditure paid for through taxation would not necessarily simply exchange private spending for government spending, because the private sector is not spending enough and is saving (i.e. the “paradox of thrift”). Like I said, I suggest that you actually map out your thoughts and go through the logic, because you are coming up with some wild arguments (that no professional Keynesian would use in their defense, to be fair).
Now, on savings and investment: Daniel, what you quoted in your first post was Austrian business cycle theory. Given that the article was written for an audience sympathizing with the Austrian business cycle theory, I see no reason to tangentially explain the Austrian theory on investment and the business cycle. In any case, it seems ridiculous that you disagree with the argument that investors will borrow more capital when said capital can be borrowed for a cheaper price. Although not explicitly referring to interest rates and capital, any macroeconomic textbook agrees that as real income rises the demand curve shifts to the right. One only needs to logically deduct that this applies to the interest rate as well (although, Keynesians seem to have a difficult time grasping what the interest rate really represent).
Apart from that, I really don’t understand what you’re asking for. You know that this is an Austrian blog, so you have a very clear idea from what perspective I will be attacking issues. The rest of your argument about savings, investment and the rate of interest does not make much sense to me, sorry.
Finally, on the topic of stimulus and time lag, professional Keynesians disagree with you (i.e. Paul Krugman; you can check older blog posts, if you’d like). The question of the time lag has been a very recent development, given that there has been a larger degree of criticism against government spending in this latest recession. Given that the effects of the stimulus remain largely unseen (well, as we all know, I believe they are non-existent, but I’ll give you the benefit of the doubt to make my point), it’s held that for the most part the effects of the stimulus are a long-term deal. Policy-makers and the Keynesian professors and theorists who support them have consistently made the argument that a considerable time-lag that should be considered (apart from any direct bailouts, which serve to pay for current expenses).
You originally said that the deficit for FY38 shouldn’t be considered, but the deficit for fiscal year FY38 ran from 1 October 1937 to 30 September 1938. As mentioned in the article, the deficit for this year was $89 million (the deficit for FY37, or the period from 1 October 1936 to 30 September 1938, was $2.189 billion, for reference). The recession began in August or September 1937 (depending on whether you cite the beginning of the market crash, or the beginning of the production crash, as the beginning of the recession) until May 1938. This all lies within FY38, and so the only deficit which is relevant is that of FY38.
You’ve clearly grossly misunderstood my point. I don’t even recognize anything I’ve ever expressed in your first paragraph, and I don’t equate a “perpetual deficit” with a “perpetual deficit growing at an accelerating pace”.
As for taxes vs. borrowing – you can’t take something I said about tax financed spending and assume it applies to debt financed spending. Taxes are taken from consumption and savings. Tax-financed spending is therefore trading off government consumption with private consumption. Debt-financed spending is taken from savings alone (not consumption). Assuming there are excessive savings (as there are now) you are NOT making the same tradeoff between private consumption and government consumption. You’re acting like the two are doing exactly the same thing.
As for ABCT – I have always held that ABCT is a necessary, but not sufficient, explanation of the business cycle.
RE: “Apart from that, I really don’t understand what you’re asking for. You know that this is an Austrian blog, so you have a very clear idea from what perspective I will be attacking issues.”
Of course – and I will critique or praise that depending on how well argued I think it is. Isn’t that the point of commenting on blogs? I understand that’s the perspective you’re coming from.
And a word of advice – it’s not very professional or classy to accuse commenters on your blog of being “inferior”, particularly commenters that enjoy reading your posts, consider them carefully, and respond at length to your points (rather than making snide remarks, like many blog commenters are inevitably reduced to).
Daniel,
I’m not sure how you are not seeing your lapse in logic. You say:
I have already shown where you contradict yourself. There is no need to repeat what you said. We are talking about the economy during a recession, and so if you agree with me that taxing and government expenditure are not counterproductive during a recessionary gap then I guess you agree with me. In that case, I am not sure why you tried to change the context in your second post. So, the failure in clarity is not with me, it’s with you (or, you are changing your mind, and acting as if that’s what you said all along). For future reference, please stay within context.
I apologize for the insult. It just seems that you are contradicting yourself, but failing to admit it.