Topic 3: The Great Depression: What caused it?
- A period of prolonged economic contraction that originated in the United States.
- 1929-1933: U.S. Real GNP and aggregate price level shrank by more than one-third, industrial output
fell by more than 50% and unemployment surged to 25% (Bordo et al., 1998:6).
- Characterised with waves of bank failure - half of all operating banks insolvent between 1929 and 1933
(Bernanke, 1983).
What caused it:
The money hypothesis: Friedman and Schwartz (1969)
- Caused by monetary contraction and inept monetary policy from the Federal Reserve.
- 1929- 1933, the money supply (M2) decreased by 35.2% (Schwartz and Friedman, 1987:27).
- This process was catalysed by widespread bank failure = impacted the money multiplier via two
channels: individuals held currency rather than bank deposits and commercial banks, worried about their
own solvency, increased their reserves in anticipation of further bank runs = decreased lending to the
real economy = money multiplier fell by a half = contracting the money supply.
- Mitchener and Richardson (2020): calculate that nine-tenths of the decline in the money multiplier can
be attributed to the decline in bank lending = debt deflation as real interest rate increased = increased
real value of debt = wave of bankruptcies = further exacerbated deflation = lower consumption and
business investment.
Non-monetary channel: Bernanke (1983)
- Complements the findings of Friedman and Schwartz (1963) but emphasises non-monetary channel.
- Reduction in money supply doesn't fully explain the crisis. Due to imperfect information, banks
differentiate between competent and incompetent borrowers via screening, monitoring and accounting.
- However, the depression impeded the ability of banks to efficiently conduct these processes = rise the
cost of credit intermediation = credit rationing = decreased aggregate demand = protracted depression as
households and firms substituted consumption with savings = paradox of thrift.
The spending hypothesis: Temin (1976)
- Criticises Friedman and Schwartz (1969) for not focusing on demand for money
- Focuses on the role of autonomous spending effects.
- Argues that banking crisis was just a continuation of issues that affected agricultural regions throughout
the 1920s - half of the bank failures between 1920-1933 happened before 1930 and many occurred
outside the time horizon studied by Friedman and Schwartz (1969) - perhaps indicating that poor
monetary policy was not responsible.
, - Temin (1976:42): precipitous fall in real consumption of 7%, between 1929 and 1930 as measured in
1958 prices - this decline is observed to be relatively larger than in other episodes of economic
contraction.
- Drop in consumption due to rapid accumulation of household debt throughout the 1920s. The stock
market crash of 1929 = sharp fall in net household wealth = increase in household leverage = fall in
consumption
- Evidence on interest rates - which hypothesis is correct? : if there was a decline in spending = shift in IS
curve = interest rate falls. If there was a decline in money supply = shift in LM curve = interest rates
rise.
- Nominal interest rates in a downtrend from 1929 onwards and throughout the banking crisis, perhaps
indicating that Friedman and Schwartz (1963) were incorrect regarding the importance of monetary
shocks (Romer and Romer, 2013). Therefore, Temin (1976) argues that the IS curve shifted more than
the LM curve = decline in spending dominated the impact of money supply changes.
- Hamilton (1987) criticises Temin. Constructs ex-post real interest rates to argue that even though
nominal rates fell, real interest rates increased = monetary policy was contractionary from 1928-31 -
supports money hypothesis theory. However, monetary policy was tighter in the 1921 recession but this
economic contraction wasn’t as severe as the depression.
Romer (1990):
- Rise in future income uncertainty due to stock market crash of 1929 caused real consumption to fall = a
potential causality between the stock market crash and the depression.
- Wealth effect channel was relatively less important as only a small fraction of society owned stocks.
This paper finds that four out of five prominent business analysts, contemporary to that of the 1920s,
became dramatically more uncertain following the Great Crash of 1929. With high uncertainty,
consumers postpone durable good consumption due to the purchase being irreversible for a long period
of time = uncertainty increases the value of waiting. A regression equation to directly test the
uncertainty hypothesis finds that for durable goods, there is a statistically significant and negative
coefficient, highlighting that large movements in stock prices depress the consumption of these goods.
- For perishable goods there is a positive coefficient, implying that stock price variability can even
stimulate the consumption of these goods, although it is statistically insignificant.
- This strengthens the argument that other factors aside from the banking crisis exacerbated the economic
crisis.
Why did Banks fail?
- Two prominent schools of thought on the role of the banking crisis: One group believes that this was
a liquidity crisis due to the Federal Reserve’s unwillingness to act as the lender of last resort.
- Opposing school of thought believes bank failures due to a solvency crisis as economy contracted and
asset prices fell, leaving little room for the Federal Reserve to intervene effectively
, Solvency crisis argument: Mason and Calomiris (2003) find the presence of ex-ante balance sheet weakness as
well as the liquidity crisis and contagion having a smaller role in the banking crisis. Argues that the Fed should
have adopted expansionary monetary policy and bank bailouts via subsidised recapitalisation. Traditional
liquidity provision may have been insufficient due to the crisis predominantly arising from solvency and
fundamental issues.
Liquidity crisis argument: on the other hand, Richardson and Troost (2009) use a natural experiment to
exploit exogenous variation in local Federal Reserve bank policy to examine the impact of differences in
monetary policy on rates of bank failure within the same state. The state of Mississippi, which was homogenous
economically and demographically, was divided into two zones under the jurisdiction of the Atlanta Fed and St.
Louis Fed respectively. The Atlanta Fed conformed to Bagehot’s rule, a doctrine that central banks should act
as a lender of last resort. In contrast, the St. Louis Fed adopted a more laissez-faire approach known as the real
bills view, where credit expansion is believed to be inflationary and comes at the expense of a slower recovery.
The St. Louis Fed therefore offered a tight discount window and had strict requirements of marginal or double
collateral, which disincentivised banks from utilising the discount window. Under the Atlanta Fed’s control,
survival rates among banks were higher, lending was more resilient, and the local economy contracted less - re-
iterates the argument made by Friedman and Schwartz (1963).
How did the depression become a global crisis?
Crafts and Fearon (2013): Gold standard acted as a fetter. Impulse to deflation came from tight U.S. monetary
policy.
- Gold standard was the mechanism which linked the world’s economies together.
- Adherence to fixed exchange rates explains timing and differential severity of crisis
- Monetary and fiscal policy were used to defend gold standard instead of alleviating falling output and
rising unemployment.
- US was a lender to the rest of the world. Gross capital inflow from the US represented 5% of German
national income. Onset of depression = US capital flow stopped and US imports declined. Germany
forced to deflate.
- Balance of payments crisis made worse by sterilisation.
- States accumulating gold were not forced to inflate their currencies. When gold losses occurred, central
banks pursued tighter monetary policy
- Between 1927 and 1932 France experienced saw its share of world gold reserves increase from 7 to 27
per cent of the total. Since the gold inflow was effectively sterilized, the policies of the Bank of France
put other countries under deflationary pressure. Irwin (2010) concludes that, on an accounting basis,
France was probably more responsible even than the US for the worldwide deflation of 1929–33. He
calculates that through their ‘gold hoarding’ policies the Federal Reserve and the Bank of France
together directly accounted for half the 30 per cent fall in prices that occurred in 1930 and 1931. This
illustrates a serious flaw in the operation of the interwar gold standard.
Tenim (1993): another channel of transmission was financial crises which spread from country to country.
- In may 1931, there was a run on Credit Anstalt (a bank in Austria) which spread to Hungary, Germany
and other European countries via 2 channels - contagion of fear and deposits by foreigners.
- A period of prolonged economic contraction that originated in the United States.
- 1929-1933: U.S. Real GNP and aggregate price level shrank by more than one-third, industrial output
fell by more than 50% and unemployment surged to 25% (Bordo et al., 1998:6).
- Characterised with waves of bank failure - half of all operating banks insolvent between 1929 and 1933
(Bernanke, 1983).
What caused it:
The money hypothesis: Friedman and Schwartz (1969)
- Caused by monetary contraction and inept monetary policy from the Federal Reserve.
- 1929- 1933, the money supply (M2) decreased by 35.2% (Schwartz and Friedman, 1987:27).
- This process was catalysed by widespread bank failure = impacted the money multiplier via two
channels: individuals held currency rather than bank deposits and commercial banks, worried about their
own solvency, increased their reserves in anticipation of further bank runs = decreased lending to the
real economy = money multiplier fell by a half = contracting the money supply.
- Mitchener and Richardson (2020): calculate that nine-tenths of the decline in the money multiplier can
be attributed to the decline in bank lending = debt deflation as real interest rate increased = increased
real value of debt = wave of bankruptcies = further exacerbated deflation = lower consumption and
business investment.
Non-monetary channel: Bernanke (1983)
- Complements the findings of Friedman and Schwartz (1963) but emphasises non-monetary channel.
- Reduction in money supply doesn't fully explain the crisis. Due to imperfect information, banks
differentiate between competent and incompetent borrowers via screening, monitoring and accounting.
- However, the depression impeded the ability of banks to efficiently conduct these processes = rise the
cost of credit intermediation = credit rationing = decreased aggregate demand = protracted depression as
households and firms substituted consumption with savings = paradox of thrift.
The spending hypothesis: Temin (1976)
- Criticises Friedman and Schwartz (1969) for not focusing on demand for money
- Focuses on the role of autonomous spending effects.
- Argues that banking crisis was just a continuation of issues that affected agricultural regions throughout
the 1920s - half of the bank failures between 1920-1933 happened before 1930 and many occurred
outside the time horizon studied by Friedman and Schwartz (1969) - perhaps indicating that poor
monetary policy was not responsible.
, - Temin (1976:42): precipitous fall in real consumption of 7%, between 1929 and 1930 as measured in
1958 prices - this decline is observed to be relatively larger than in other episodes of economic
contraction.
- Drop in consumption due to rapid accumulation of household debt throughout the 1920s. The stock
market crash of 1929 = sharp fall in net household wealth = increase in household leverage = fall in
consumption
- Evidence on interest rates - which hypothesis is correct? : if there was a decline in spending = shift in IS
curve = interest rate falls. If there was a decline in money supply = shift in LM curve = interest rates
rise.
- Nominal interest rates in a downtrend from 1929 onwards and throughout the banking crisis, perhaps
indicating that Friedman and Schwartz (1963) were incorrect regarding the importance of monetary
shocks (Romer and Romer, 2013). Therefore, Temin (1976) argues that the IS curve shifted more than
the LM curve = decline in spending dominated the impact of money supply changes.
- Hamilton (1987) criticises Temin. Constructs ex-post real interest rates to argue that even though
nominal rates fell, real interest rates increased = monetary policy was contractionary from 1928-31 -
supports money hypothesis theory. However, monetary policy was tighter in the 1921 recession but this
economic contraction wasn’t as severe as the depression.
Romer (1990):
- Rise in future income uncertainty due to stock market crash of 1929 caused real consumption to fall = a
potential causality between the stock market crash and the depression.
- Wealth effect channel was relatively less important as only a small fraction of society owned stocks.
This paper finds that four out of five prominent business analysts, contemporary to that of the 1920s,
became dramatically more uncertain following the Great Crash of 1929. With high uncertainty,
consumers postpone durable good consumption due to the purchase being irreversible for a long period
of time = uncertainty increases the value of waiting. A regression equation to directly test the
uncertainty hypothesis finds that for durable goods, there is a statistically significant and negative
coefficient, highlighting that large movements in stock prices depress the consumption of these goods.
- For perishable goods there is a positive coefficient, implying that stock price variability can even
stimulate the consumption of these goods, although it is statistically insignificant.
- This strengthens the argument that other factors aside from the banking crisis exacerbated the economic
crisis.
Why did Banks fail?
- Two prominent schools of thought on the role of the banking crisis: One group believes that this was
a liquidity crisis due to the Federal Reserve’s unwillingness to act as the lender of last resort.
- Opposing school of thought believes bank failures due to a solvency crisis as economy contracted and
asset prices fell, leaving little room for the Federal Reserve to intervene effectively
, Solvency crisis argument: Mason and Calomiris (2003) find the presence of ex-ante balance sheet weakness as
well as the liquidity crisis and contagion having a smaller role in the banking crisis. Argues that the Fed should
have adopted expansionary monetary policy and bank bailouts via subsidised recapitalisation. Traditional
liquidity provision may have been insufficient due to the crisis predominantly arising from solvency and
fundamental issues.
Liquidity crisis argument: on the other hand, Richardson and Troost (2009) use a natural experiment to
exploit exogenous variation in local Federal Reserve bank policy to examine the impact of differences in
monetary policy on rates of bank failure within the same state. The state of Mississippi, which was homogenous
economically and demographically, was divided into two zones under the jurisdiction of the Atlanta Fed and St.
Louis Fed respectively. The Atlanta Fed conformed to Bagehot’s rule, a doctrine that central banks should act
as a lender of last resort. In contrast, the St. Louis Fed adopted a more laissez-faire approach known as the real
bills view, where credit expansion is believed to be inflationary and comes at the expense of a slower recovery.
The St. Louis Fed therefore offered a tight discount window and had strict requirements of marginal or double
collateral, which disincentivised banks from utilising the discount window. Under the Atlanta Fed’s control,
survival rates among banks were higher, lending was more resilient, and the local economy contracted less - re-
iterates the argument made by Friedman and Schwartz (1963).
How did the depression become a global crisis?
Crafts and Fearon (2013): Gold standard acted as a fetter. Impulse to deflation came from tight U.S. monetary
policy.
- Gold standard was the mechanism which linked the world’s economies together.
- Adherence to fixed exchange rates explains timing and differential severity of crisis
- Monetary and fiscal policy were used to defend gold standard instead of alleviating falling output and
rising unemployment.
- US was a lender to the rest of the world. Gross capital inflow from the US represented 5% of German
national income. Onset of depression = US capital flow stopped and US imports declined. Germany
forced to deflate.
- Balance of payments crisis made worse by sterilisation.
- States accumulating gold were not forced to inflate their currencies. When gold losses occurred, central
banks pursued tighter monetary policy
- Between 1927 and 1932 France experienced saw its share of world gold reserves increase from 7 to 27
per cent of the total. Since the gold inflow was effectively sterilized, the policies of the Bank of France
put other countries under deflationary pressure. Irwin (2010) concludes that, on an accounting basis,
France was probably more responsible even than the US for the worldwide deflation of 1929–33. He
calculates that through their ‘gold hoarding’ policies the Federal Reserve and the Bank of France
together directly accounted for half the 30 per cent fall in prices that occurred in 1930 and 1931. This
illustrates a serious flaw in the operation of the interwar gold standard.
Tenim (1993): another channel of transmission was financial crises which spread from country to country.
- In may 1931, there was a run on Credit Anstalt (a bank in Austria) which spread to Hungary, Germany
and other European countries via 2 channels - contagion of fear and deposits by foreigners.
