Prof. Bryan Caplan

Econ 816

Spring, 2000


Week 13: The Great Depression and European Unemployment

I.                  The Monetary Explanation Proposed: The Classic Case of Friedman and Schwartz

A.                 As Bernanke points out, the Great Depression can be seen as two puzzles:

1.                  The AD puzzle: Why did AD suddenly fall?

2.                  The AS puzzle: Why did LRAS take so long to restore full employment at lower level of demand?

B.                 The AD puzzle has gotten much more attention. 

C.                Most prominent solution to AD puzzle pioneered by Friedman and Schwartz (1963), unseating the dinosaur Keynesian fiscalist orthodoxy.

1.                  Interesting footnote: As Selgin and White point out in the article on privatization of money, one major factor in the later bank runs was the fear that the U.S. would abandon the gold standard.

D.                The monetary facts for the U.S. (you've seen them before):



% ch


% ch




































E.                 Bordo, Choudri, and Schwartz ("Could Stable Money Have Averted the Great Contraction") provides extremely interesting - if speculative - evidence that monetary factors were decisive.  After estimating models of the U.S. economy using interwar U.S. data, they simulate alternative policy rules and predictably find that stable money would have either greatly moderated the depression or eliminated it entirely.

II.               The Monetary Interpretation Confirmed: The International Evidence of Eichengreen and Bernanke

A.                 Note Bernanke's identity: M1=(M1/Base)*(Base/Res)*(Res/Gold)*Pgold*Qgold, where:

1.                  M1/Base is the money multiplier.

2.                  Base/Res is the inverse of the gold backing ratio.

3.                  Res/Gold is the ratio of international reserves to gold.

B.                 The CGS vs. the inter-war "gold-exchange" standard: The gold standard was only very briefly restored during the inter-war period.  Moreover, this restoration was actually a large modification - and weakening - of the CGS:

1.                  Switch from gold coin to gold bullion to curtail domestic demand (e.g. 400 oz. minimum in UK).

2.                  Peripheral and some core countries switch substantially from gold to foreign exchange reserves.  Under CGS Res/Gold=1; under gold-exchange standard Res/Gold>1. 

3.                  This left the international monetary system vulnerable to an international run that could not have happened under the CGS.

C.                The main reason for the regime change: wartime money creation was enormous, and thus required massive deflation to return at par in an orthodox way.  The switch to the "mixed" gold-exchange system and eliminate non-bullion redemption was an attempt to return to the gold standard without devaluation OR severe deflation.

D.                Bernanke's AD findings:

1.                  Evidence of large regime differences (between on-gold and off-gold countries) from panel data:

a)                 Manufacturing production higher for non-gold countries.

b)                 Price growth higher for non-gold.

c)                  M1 growth higher for non-gold.

d)                 Nominal wages fall less in non-gold.

e)                 Real wages fall more in non-gold.

f)                    Employment rises more in non-gold.

g)                 Nominal interest rates a little higher in non-gold.

h)                  Ex post real rates lower in non-gold (except last year).

2.                  Note Bernanke's tabulations of the factors contributing to decline in the money supply.  (Note further that U.S. basically lost no gold at any time!)

E.                 In general: as Bernanke points out, international comparisons are extremely helpful, and the ideological factors behind monetary regimes strongly indicate that the connection between money and the depressions are causal.

III.            The AS Puzzle: Whatever Happened to Long-Run Neutrality?

A.                 Monetary forces explain the severity of the nominal decline, but it seems necessary to look at labor market policies and other interventions to explain why the real reaction to the nominal decline was so sharp.  Earlier depressions exhibited short-run nominal-real connections, but swift recoveries.  The Great Depression often seemed to have stable high unemployment in the absence of expansionary monetary policy.

B.                 Rothbard's main contribution: debunking "the myth of laissez-faire."  Herbert Hoover was the most extreme interventionist to occupy the White House prior to FDR.  A few interesting pieces of evidence.

1.                  FDR supported Hoover's presidential bid in 1920.

2.                  Hoover unsuccessfully pushed for federal anti-depression policy in 1920-21.

3.                  Hoover repeatedly indicated that he believed that low wages cause unemployment, apparently on the basis of nonsensical "buy-back-the-product"-type arguments.

4.                  Hoover quote, 1932 presidential campaign (p.127)

C.                What were Hoover's actual anti-depression policies like?

1.                  Major conferences held, in which Hoover elicited pledges from major industrial leaders not to cut nominal wages.  (Why no pledge to avoid lay-offs?)

2.                  Federal spending increases from $3.3 billion in 1929 to $4.6 billion in 1932 in nominal terms.  (3.1% S/GDP to 7.9% S/GDP).  Large increase in public works spending.

3.                  Rothbard actually attacks the "inflationist" policies of Hoover.  In terms of the monetary base, this is not wrong.

4.                  Reconstruction Finance Corporation created, partly in order to encourage sound banks to help shaky ones.  $2.3 billion in loans in 1932.

5.                  Federal Farm Board to cartelize agriculture actually pushed through by Hoover in early 1929 before the depression even began: $500 million revolving fund for price support.  Government purchases try to hold up agricultural prices once depression hits, but these fail due to lack of compulsory quantity limits.

6.                  Why does anyone believe that Hoover followed laissez-faire policies then?  Based almost entirely on Hoover's opposition to the dole (note that evidence of negative consequences on British economy were already evident), and the fact that FDR took the policies considerably further.

D.                Discarding the myth of Hoover's policies permits a more general study of labor market and other interventions during the Great Depression.  Internationally, these were ubiquitous, and provide a good explanation for the actual rise in real wages during the Great Depression (with the rise continuing through 1935 in on-gold countries).

E.                 Bernanke's evidence:

1.                  Regressing output on nominal wage and price shows that higher wages reduce output, and higher prices increase output (although note hard to reject hypothesis of real rigidity when tested against hypothesis of nominal rigidity).

2.                  Nominal wages vs. output in gold bloc countries.

F.                 Interesting case studies:

1.                  U.S. labor market and industrial policy under Roosevelt. (Schlesinger v.1, p.249; v.2, p.385)

2.                  Germany (Barkai, pp.250-1, 256-7)

IV.            European Unemployment and Real Rigidities

A.                 Eichengreen and Bernanke note that France was an outlier in the Great Depression: even after France adopted expansionary monetary policy in 1936, it seemed unable to reduce unemployment.  Rather, workers just forced up nominal wages to match inflation.

B.                 Since the '70's, the same sort of condition has become widespread throughout most of Europe.  Basic facts:

1.                  Visual inspection: While U.S. unemployment rate has ups and downs, European unemployment rises, then plateaus, then rises again - with only one brief recovery.

2.                  Total employment increase from 1970-1996: 58% for U.S., 12% for Europe (47 M vs. 18 M; note European population is greater).

3.                  Labor force participation: in the U.S., rise from 65% to 75%; in Europe, fall from 65% to 60%.

C.                What explanations can be ruled out?

1.                  Oil shocks

2.                  Productivity growth decline

3.                  Also: Earlier argument about the U-shaped corporativism curve weakened by drastic rise of Scandinavian unemployment from 1990.

D.                Unemployment is permanently high, as in the Great Depression.  Supply shocks don't seem important, so the labor market seems to be the right place to look.

E.                 Like France after 1936, expansionary policy seems impotent to solve the problem.  The natural theoretical framework to turn to understand this is thus not nominal rigidities, but real rigidities.

V.               Exacerbating Real Rigidities: European Unemployment Policies

A.                 While real rigidity theories have been generally designed to help understand unemployment in less regulated economies, labor market regulation can make the problem a lot worse - and this is probably what has happened in Europe.  (See Nickell, Tables 4&5).

B.                 High legal minimum wages.  (E.g. 34% of median in U.S. vs. 60% in France).

C.                High unemployment/welfare benefits with long durations.

D.                Firing/layoff regulations.

E.                 Mandatory benefits (vacation, sick leave, maternity leave, etc.)  (How does the interaction between mandatory benefits and nominal and real rigidity work?)

F.                 High unionization rates with strong legal support for unions.  (Note: In some countries like France, non-union workers still have their wages determined by union negotiations.)

VI.            Arguments Against the Obvious Explanation

A.                 Most of the interventions existed for at least a decade before they had any observable impact.  During the '70's European unemployment was lower than that in the U.S.  Running panel data regressions with time dummies would probably show little impact of labor market policies.

B.                 Need to account for pre-existing labor market differences.  The standard way to do this is to try correlate changes in policy with changes in unemployment.  If you do this with panel data, it is known as "differences-in-differences" estimation, and it too tends to show little impact.  Or you could just do panel data regressions with country dummies; the high-unemployment countries with have large positive country dummies, but would not be too likely to show a big impact of regulation/labor market rigidity.

C.                Arguments from Nickell:

1.                  Looking at decomposed European countries unemployment rates changes the picture: often the more rigid countries have lower unemployment.

2.                  Need to control for demographics, cultural factors, etc.

3.                  High replacement rate not a problem so long as the duration of benefits is reasonable.

4.                  Unionization not a problem so long as it is offset by coordination between employers and unions.

5.                  High taxation combined with high minimum wages is a problem, but strict employment legislation is not.

D.                Rebuttals:

1.                  Explanation for delayed effect: unemployment caused by interaction between productivity/supply shocks and labor market rigidities.  Makes use of time dummies inappropriate.

2.                  Also: much of the impact may be long-term.  Developed human capital will be used, but the incentive to develop new human capital falls.  Same goes for "cultural capital" like the work ethic.  Makes use of country dummies at least suspect.

3.                  Scandinavian and UK outliers have disappeared.

4.                  Others?

E.                 The moral: Comparing European unemployment to the Great Depression shows the danger of over-stating the importance of nominal factors - and the ease of nominal solutions.