Real Origins of the Great Depression - uni-heidelberg.de Origins of the Great Depression Monopoly...
Transcript of Real Origins of the Great Depression - uni-heidelberg.de Origins of the Great Depression Monopoly...
Real Origins of the Great DepressionMonopoly Power, Unions and the Interwar Business Cycle
Monique Ebell Albrecht Ritschl
HU Berlin & LSE LSE & CEPR
May 2010
Ebell/Ritschl (LSE & HU Berlin) Real Origins of the Great Depression May 2010 1 / 43
Previous Work on the Interwar Economy
Business Cycle Accounting approach of Chari, Kehoe & McGrattan(2002)
Key role for effi ciency and labor wedgesFocus on Great Depression and aftermath 1929-39Cole & Ohanian (2002) for UK, Hornstein (2002) for Germany,Beaudry & Portiers (2002) for France
Business Cycle Accounting framework
Perfect compeition in goods marketsWalrasian labor markets
Cole and Ohanian (2004)
Explicit modelling of frictionsSpecific to New Deal institutionsFocus on slow recovery from Great Depression 1933 onwards
Ebell/Ritschl (LSE & HU Berlin) Real Origins of the Great Depression May 2010 2 / 43
Our Contributions
1 Scope
Severe recession 1919-21, accompanying stock market crashRecovery to trend 1922-29, stock market boomOnset of the Great Depression 1929-31, stock market crash
2 Explicit modelling of frictions
Interactions between monopoly power and wage bargainingUnemploymentExplanatory power for asset markets
3 History
Identify 3 major shifts in legal environment for organized labor:1914/19, late 1921, 1929Threat point of organized labor strengthened in a series ofdevelopments during 1929, culminating in October 1929
Ebell/Ritschl (LSE & HU Berlin) Real Origins of the Great Depression May 2010 3 / 43
Real Dow Jones Industrial Average, 1913-1939Relative to 1.8% trend
1913 1914 1916 1918 1919 1920 1921 1922 1923 1924 1925 1926 1927 1928 1929 1930 1931 1932 1933 1934 1935 1936 1937 1938 19390.0
50.0
100.0
150.0
200.0
250.0
300.0
Ebell/Ritschl (LSE & HU Berlin) Real Origins of the Great Depression May 2010 4 / 43
Real Private Non-Farm Per Capita GDPRelative to 1.8% trend
1913 1914 1915 1916 1917 1918 1919 1920 1921 1922 1923 1924 1925 1926 1927 1928 1929 1930 1931 1932 1933 1934 1935 1936 1937 1938 193940
50
60
70
80
90
100
110
Ebell/Ritschl (LSE & HU Berlin) Real Origins of the Great Depression May 2010 5 / 43
The Interwar Business Cycle
NonFarm Unemployment 19131939
0
5
10
15
20
25
30
35
1913
1914
1915
1916
1917
1918
1919
1920
1921
1922
1923
1924
1925
1926
1927
1928
1929
1930
1931
1932
1933
1934
1935
1936
1937
1938
1939
Ebell/Ritschl (LSE & HU Berlin) Real Origins of the Great Depression May 2010 6 / 43
Real Wages in Manufacturing, 1929-1939Relative to 1.8% trend, deflated by PPI
1928 1929 1930 1931 1932 1933 1934 1935 1936 1937 1938 1939
0.900
0.950
1.000
1.050
1.100
1.150
1.200
Ebell/Ritschl (LSE & HU Berlin) Real Origins of the Great Depression May 2010 7 / 43
Interwar Labor Law Regimes
1 Clayton Act 1914-21Clayton Act of 1914 exempts unions from anti-trust legislationTrade union activity rises sharply after WWI
2 Laissez Faire 1922-292 Supreme Court rulings in December 1921 invalidate the pro-unionprovisions of the Clayton ActPicketing prohibited and subject to injunctionTrade union activity plummetsLaissez-faire anti-trust policy of Coolidge administration increasesmonopoly power
3 Welfare Capitalism 1929-32Texas & N.O. Railroad case [DC August 1928, CC Appeals June 1929,Supreme Court May 1930] marks a sea change in labor lawReversal of a unanimous district and appellate court decision wouldhave been highly unusual, marking a break with precedentIncreasing pro-union legislative efforts in summer/fall 1929High wage policy of Hoover
Ebell/Ritschl (LSE & HU Berlin) Real Origins of the Great Depression May 2010 8 / 43
Texas & N.O. and Pro-Union Legislation
January - March 1929Shipstead anti-injunction bill introduced and quickly tabled by theSenate Judiciary committee due to fears of unconstitutionality.
June 19292nd Circuit Court of Appeals upholds pro-union Texas & N.O. ruling.
August 1929Delegation from Senate Judiciary Committee meets with executivecouncil of the AFL to propose ’robust’anti-injunction legislation.
October 14 -18, 1929AFL national convention in Toronto, Senate proposal adopted.
October 21 - 29, 1929First weakness in stock market, followed by Black Thursday, BlackMonday, Black Tuesday
Ebell/Ritschl (LSE & HU Berlin) Real Origins of the Great Depression May 2010 9 / 43
Contemporary Views
In an article published in the AER 12/30, Edward Berman writes:
“The Supreme Court’s decision in the Texas and N.O.Railroad case is without doubt one of the most importantrendered in a labor case in many years. Considered as a whole, itmay be regarded as a great victory for organized labor. . . . Itputs the Supreme Court on record in favor of legislation designedto promote collective bargaining. It promises that the court will,in the future, be more friendly to state and federal legislationdesigned to protect workers from the coercive activities ofanti-union employers. ”
Ohanian emphasizes Hoover’s role in persuading leading industrialiststo implement high wage policy.Threat of re-unionization if firms did not offer high wages.Proto-Keynesian idea that high wages would prop up consumption,averting a "disastrous" recession as in 1920-21.
Ebell/Ritschl (LSE & HU Berlin) Real Origins of the Great Depression May 2010 10 / 43
NY Times, October 16, 1929
Ebell/Ritschl (LSE & HU Berlin) Real Origins of the Great Depression May 2010 11 / 43
NY Times, October 27, 1929
Ebell/Ritschl (LSE & HU Berlin) Real Origins of the Great Depression May 2010 12 / 43
Model
Monopolistic competition in the goods market
Mortensen-Pissarides search frictions in the labor market
Workers choose to search among firms of type k ∈ {I ,C}Individual =⇒ each worker is marginal workerCollective =⇒ workers can garner share of profits
Complete asset markets, risk-neutral agents
Productivity growth constant and deterministic
No capital
Ebell/Ritschl (LSE & HU Berlin) Real Origins of the Great Depression May 2010 13 / 43
Agents as Consumers
Risk neutral, no savings
Dixit-Stiglitz preferences over continuum of differentiated goods
Elasticity of substitution across goods −σ
Demand function facing firm i :
yi =(p (yi )P
)−σ
Y
Ebell/Ritschl (LSE & HU Berlin) Real Origins of the Great Depression May 2010 14 / 43
Labor Market Search Frictions
Segmented labor markets k ∈ {C , I}Matching function
m (Uk ,Vk ) = mUηkV
1−ηk
Job-finding rate
m (Uk ,Vk )Uk
= mθ1−ηk = f (θk )
Job-filling ratem (Uk ,Vk )
Vk= mθ
−ηk = q (θk )
where labor market tightness θk ≡ VkUk.
Beveridge CurveUk · f (θk ) = (1− Uk ) χ
Uk =χ
χ+ f (θk )
Ebell/Ritschl (LSE & HU Berlin) Real Origins of the Great Depression May 2010 15 / 43
Agents as Workers
Value of unemployment in segment k ∈ {I ,C}
V Uk = b+1
1+ r
[fkV
E ′k + (1− fk )V U ′k
]Value of employment
V Ek = wk +1
1+ r
[(1− χ)V E ′k + χV U ′k
]where χ is the exogenous separation rate.
Worker’s surplus
VWk = wk − b+ β (1− χ− fk )VW ′k
Ebell/Ritschl (LSE & HU Berlin) Real Origins of the Great Depression May 2010 16 / 43
Firms
Value of the firm
V Jk (hk ) = maxvk
{p (yk )P
yk − wkhk − κvk +1
1+ rV J(h′k)}
subject to
demand :p (yk )P
=(ykY
)− 1σ
technology : yk = Ahktransition: h′k = (1− χ) hk + q (θk ) vkwage curve : wk = wk (hk )
Firm’s Euler Equation (labor demand)
κ
q (θk )=
11+ r
(σ− 1
σAp(y ′k)− w ′k − h′k
∂w ′k∂h′k
+ (1− χ)κ
q(θ′k))
Ebell/Ritschl (LSE & HU Berlin) Real Origins of the Great Depression May 2010 17 / 43
Wage Bargaining Regimes
Individual Bargaining
Each worker is treated as marginal workerWorker’s threat point: Firm loses marginal workerBargained wage includes a share of hiring costs
Right-to-Manage Collective Bargaining
Workers form a bargaining coalitionWorkers’threat point: Firm shuts downBargained wage includes a share of profits
Ebell/Ritschl (LSE & HU Berlin) Real Origins of the Great Depression May 2010 18 / 43
Individual Wage Bargaining
Nash bargaining problem
maxwI
β lnVWI + (1− β) ln∂V JI (hI )
∂hI
IB wage curve
wI = (1− β)r
1+ rV U + β
σ− 1σ− β
Ap (yI )P
with slope∂wI∂hI
= −β
σ
σ− 1σ− β
Ap (y)
Ebell/Ritschl (LSE & HU Berlin) Real Origins of the Great Depression May 2010 19 / 43
Firm’s Euler equation at the steady state (labor demand)
wI =σ− 1σ− β
Ap (yI )P− (r + χ)
κ
q (θI )
Hiring externality (overhiring), interaction between σ and β
wI + (r + χ)κ
q (θI )︸ ︷︷ ︸employment costs
=σ
σ− β︸ ︷︷ ︸>1
σ− 1σ
p (yI )P︸ ︷︷ ︸
<1
A︸︷︷︸MPL︸ ︷︷ ︸
MRP
Hire workers beyond the point at which employment costs covered byMRP.
Why? Adding a worker depresses the wages of all workers. [Stole andZwiebel (AER, REStud 1996)]
Ebell/Ritschl (LSE & HU Berlin) Real Origins of the Great Depression May 2010 20 / 43
Right-to-Manage Collective Bargaining
Right-to-Manage bargaining problem
maxwC
β ln(hCV
WC
)+ (1− β) lnV JC
CB wage curve
wC −rV UC1+ r︸ ︷︷ ︸
worker’s surplus
=β
1− β
[Ap (yC )P− wC − χ
κ
q (θC )
]︸ ︷︷ ︸
profit per worker
CB labor demand
wC =σ− (1− β)
σAp (yC )− (r + χ)
κ
q (θC )
Ebell/Ritschl (LSE & HU Berlin) Real Origins of the Great Depression May 2010 21 / 43
Bargained Wages
Individual Bargaining
wI =rV UI1+ r
+β
1− β(r + χ)
κ
q (θI )
Collective Bargaining
wC =rV UC1+ r
+β
σ− 1
[rV UC1+ r
+κ
q (θC )
(r
1− βσ+ χ
)]
Ebell/Ritschl (LSE & HU Berlin) Real Origins of the Great Depression May 2010 22 / 43
Equilibrium
Market Clearing when share µ ∈ [0, 1] of CB firms
I = (1− µ)p (yI )P
yI + µp (yC )P
yC
Segmented labor markets, workers indifferent between searching in Ior C market
rV UI1+ r
=rV UC1+ r
—> 2 equilibrium conditions in the two unknowns (θI , θC ).
Once have equilibrium tightnesses, can easily obtain remainingvariables.
Ebell/Ritschl (LSE & HU Berlin) Real Origins of the Great Depression May 2010 23 / 43
Quantitative Results
Parameter Value TargetA 1.0 normalizationr 0.01 4% annual rη 0.50 empirical estimatesβ 0.50 Hosios conditionb/w 0.30 post-war b/w ≥ 0.40
χ 0.118 BLS dataκ 0.19 perf. comp. u = 5.5%
—> all equilibrium variables as functions of (µ, σ)
Ebell/Ritschl (LSE & HU Berlin) Real Origins of the Great Depression May 2010 24 / 43
0 5 10 15 200.5
0.6
0.7
0.8
0.9
1Output
0 5 10 15 20
0.65
0.7
0.75
0.8
0.85
0.9
0.95
1
Competition σ
Wages
0 5 10 15 200
0.05
0.1
0.15
0.2
0.25
0.3
0.35
Competition σ
Profits
0 5 10 15 200
0.1
0.2
0.3
0.4
0.5Unemployment
IndividualCollective
—> need to pin down time series of µ and σ
Ebell/Ritschl (LSE & HU Berlin) Real Origins of the Great Depression May 2010 25 / 43
Mechanism: Shift towards Collective Bargaining
Collective bargaining wage contains a share of monopoly rents
Total surplus depends on profits, so via wage bargaining, workers’andfirms’incentives to maximize profits are well-alignedBy first principles, monopolistic(ally competitive) profits are maximizedby restricting output below the socially effi cient level.
Under collective bargaining, workers take a share of profits roughlyequal to µ —> asset values decline
Real wages increase, employment decreases —> output decreases
Vacancies decrease, tightness increases —> unemployment rises
Ebell/Ritschl (LSE & HU Berlin) Real Origins of the Great Depression May 2010 26 / 43
Baseline Scenario
Year σ Markup σ−βσ−1 µ
1913-1918 4.0 16.7% 01919 4.0 16.7% 0.251920 4.0 16.7% 0.501921 4.0 16.7% 0.751922-1924 4.0 16.7% 01925-1926 3.5 20% 01927 3.0 25% 01928 2.5 33.3% 01929 2.3 38.5% 01930 2.3 38.5% 0.251931-39 2.3 38.5% 0.50
—> σ1913−21 so that ∆µ accounts for 60% of drop in DJ Industrials—> σ1929 so that ∆µ plus increase in σ accounts for 60% of increase in DJIndustrialsEbell/Ritschl (LSE & HU Berlin) Real Origins of the Great Depression May 2010 27 / 43
Asset Prices
1910 1915 1920 1925 1930 1935 1940
60
80
100
120
140
160
180
200
220
240
260Dow Jones: Baseline Scenario
data 1.8pc trendmodel
Ebell/Ritschl (LSE & HU Berlin) Real Origins of the Great Depression May 2010 28 / 43
Private Non-Farm GDP
1910 1915 1920 1925 1930 1935 194050
60
70
80
90
100
110Nonfarm Private GDP: Baseline Scenario
data 1.8 pc trendmodel
Ebell/Ritschl (LSE & HU Berlin) Real Origins of the Great Depression May 2010 29 / 43
Unemployment
1910 1915 1920 1925 1930 1935 19400
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4Unemployment: Baseline Scenario
data nonfarm misplaceddata nonfarmmodel
Ebell/Ritschl (LSE & HU Berlin) Real Origins of the Great Depression May 2010 30 / 43
Wages
1929 1930 1931 1932 1933 1934 1935 1936 1937 1938 193995
100
105
110
115
120Wages: Baseline Scenario
datamodel
Ebell/Ritschl (LSE & HU Berlin) Real Origins of the Great Depression May 2010 31 / 43
Maximal Scenario
1910 1915 1920 1925 1930 1935 1940
60
80
100
120
140
160
180
200
220
240
260Dow Jones: Maximal Scenario
data 1.8% trendmodel
Ebell/Ritschl (LSE & HU Berlin) Real Origins of the Great Depression May 2010 32 / 43
Maximal Scenario
1910 1915 1920 1925 1930 1935 194050
60
70
80
90
100
110Nonfarm Private GDP: Maximal Scenario
data 1.8% trendmodel
Ebell/Ritschl (LSE & HU Berlin) Real Origins of the Great Depression May 2010 33 / 43
Maximal Scenario
1910 1915 1920 1925 1930 1935 19400
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
0.45Unemployment: Maximal Scenario
data misplaceddatamodel
Ebell/Ritschl (LSE & HU Berlin) Real Origins of the Great Depression May 2010 34 / 43
Maximal Scenario
1929 1930 1931 1932 1933 1934 1935 1936 1937 1938 193995
100
105
110
115
120Wages: Maximal Scenario
datamodel
Ebell/Ritschl (LSE & HU Berlin) Real Origins of the Great Depression May 2010 35 / 43
A Counterfactual
What if Hoover had vigorously enforced anti-trust, reducing thedegree of monopoly power in the economy?
Asset prices would have declinedIncentives to unionize would have been diminishedIf switch to CB: less severe recesssionIf stay with IB: no recession, slight expansionary effect
Ebell/Ritschl (LSE & HU Berlin) Real Origins of the Great Depression May 2010 36 / 43
Summary
1 Model accounts well for
Early 1920s recession & stock market crashRoaring 1920s expansion & stock market boomOnset of the Great Depression 1929-31
2 Interactions between monopoly power and wage bargaining3 Crucial difference between individual and collective bargaining is thethreat point
Measures of union membership either overstate or understateimportance of collective bargainingLegal environment more importantHanes (2000, AER) finds that time it took for firms to reduce wagesafter 1929 (during deflation) positively correlated with sectoral productmarket concentration
Ebell/Ritschl (LSE & HU Berlin) Real Origins of the Great Depression May 2010 37 / 43
Individual Wage Bargaining - Details
Nash bargaining problem
maxwI
µ lnVWI + (1− µ) ln∂V JI (hI )
∂hI
subject to
∂V JI (hI )∂hI
=σ− 1
σAp (y)− wI − hI
∂wI∂hI
+ (1− χ)Φv
q (θI )
VWI = wI − b+ β (1− χ− fI )VW ′I
FOC
wI = (1− µ) b+µ
[σ− 1
σAp (y)− hI
∂wI∂hI
+ (1− χ)Φv
q (θI )
]− (1− µ) β (1− χ− fI )VW ′I
(1)Conjecture: VW ′I is independent of hI and wI . Solve first-order lineardifferential equation (1), then confirm conjecture.
Ebell/Ritschl (LSE & HU Berlin) Real Origins of the Great Depression May 2010 38 / 43
Individual Wage Bargaining - Details
FOLDE
wI = (1− µ) b+µ
[σ− 1
σAp (y)− hI
∂wI∂hI
+ (1− χ)Φv
q (θI )
]− (1− µ) β (1− χ− fI )VW ′I
Solution to FOLDE
wI = (1− µ) b+µ
[σ− 1σ− µ
Ap (y) + (1− χ)Φv
q (θI )
]− (1− µ)
11+ r
(1− χ− fI )VW ′I
(2)with
hI∂wI∂hI
= −µ
σ
σ− 1σ− µ
Ap (y) (3)
(3) into firm’s Euler equation
Φv
q (θ)=
11+ r
(σ− 1σ− µ
Ap(y ′)− w ′ + (1− χ)
Φv
q(θ′))
Solve (2) for VW ′I , substitute back into worker’s surplus
VWI =µ
1− µ
(σ− 1σ− µ
Ap (y)− wI + (1− χ)Φv
q (θI )
)Ebell/Ritschl (LSE & HU Berlin) Real Origins of the Great Depression May 2010 39 / 43
Individual Wage Bargaining - Details
(3) into firm’s Euler equation
Φv
q (θ)=
11+ r
(σ− 1σ− µ
Ap(y ′)− w ′ + (1− χ)
Φv
q(θ′))
Solve (2) for VW ′I , substitute back into worker’s surplus
VWI =µ
1− µ
(σ− 1σ− µ
Ap (y)− wI + (1− χ)Φv
q (θI )
)One step ahead
11+ r
VW ′I =µ
1− µ
11+ r
(σ− 1σ− µ
Ap(y ′)− w ′I + (1− χ)
Φv
q(θ′I))︸ ︷︷ ︸
Φv/q(θI )
Future surplus depends only on aggregate variables. Search costsdepend only on aggregate variables.Wage curve
wI = (1− µ) b+ µ
[σ− 1σ− µ
Ap (y) +Φv θI
]Ebell/Ritschl (LSE & HU Berlin) Real Origins of the Great Depression May 2010 40 / 43
Equilibrium Wages
Individual Bargaining
wI =rV UI1+ r
+β
1− β(r + χ)
κ
q (θI )
Collective Bargaining
wC =rV UC1+ r
+β
σ− 1
[rV UC1+ r
+κ
q (θC )
(r
1− βσ+ χ
)]
Ebell/Ritschl (LSE & HU Berlin) Real Origins of the Great Depression May 2010 41 / 43
Equilibrium I
Market Clearing, share µ ∈ [0, 1] of CB firms
I = (1− µ)p (yI )P
yI + µp (yC )P
yC
Yields the first equilibrium condition relating θI and θC
A1−σ = (1− µ)
{σ− β
σ− 1
[rV UI1+ r
+1
1− β(r + χ)
κ
q (θI )
]}1−σ
+µ
{σ
σ− 1
[rV UC1+ r
+
(1
1− βr + χ
)κ
q (θC )
]}1−σ
Ebell/Ritschl (LSE & HU Berlin) Real Origins of the Great Depression May 2010 42 / 43
Equilibrium II
Segmented labor markets, workers indifferent between searching in Ior C market
rV UI1+ r
=rV UC1+ r
Yields a second condition relating θI and θC
b+β
1− βκθI
=r + χ
r + χ− f (θC ) βσ−1
b+f (θC )
βσ−1
r + χ− f (θC ) βσ−1
κ
q (θC )
(r
1− βσ+ χ
)Equilibrium (θI , θC ) pinned down for each (µ, σ) pair.
Ebell/Ritschl (LSE & HU Berlin) Real Origins of the Great Depression May 2010 43 / 43