Post on 18-Feb-2021
The Rise of Market Power
Jan Eeckhout
UPF Barcelona and UCL
BCCP ConferenceJune 21, 2019
Facts about Market Power
Estimating markups
• Cost based method; publicly traded firms 1955–2016• From firm’s FOC for cost minimization and µ = PMC :
µit = θVit
PitQit
PVit VitV ∈ {Lab,Mat,Elec , ...}
• Individual Markup ⇒ distribution of markups• Average markup, weighted by mit (sales, costs, employment,...):
µt =∑i
mitµit
• Markup 6= Market Power: with fixed cost calculate profit rate
1. HeterogeneityNo Change... in median markup
1.1
1.2
1.3
1.4
1.5
1.6
1960 1970 1980 1990 2000 2010
1. HeterogeneityIncrease in Average Markup since 1980
1.1
1.2
1.3
1.4
1.5
1.6
1960 1970 1980 1990 2000 2010
AverageP50
1. HeterogeneityAll Action in Upper Half Distribution
1
1.5
2
2.5
1960 1970 1980 1990 2000 2010
AverageP90P75P50
1. HeterogeneityKernel Density 1980, 2016
0
.5
1
1.5
2
1 2 3
20161980
Facts
1. Heterogeneity: sharp rise for few firms; no rise for most
(Carlos Brito)
Reallocation of sales from low to high markup firms (2/3)
Technology Matters:
Fixed cost (SG&A) ↑Returns to Scale ↑
Magnitude of the Increase?
Weighting and Aggregation is crucial (+15-20 points)Profit rate (+7-8 ppts) 6= Markup (+30-40 points)
Facts
1. Heterogeneity: sharp rise for few firms; no rise for most (Carlos Brito)
Reallocation of sales from low to high markup firms (2/3)
Technology Matters:
Fixed cost (SG&A) ↑Returns to Scale ↑
Magnitude of the Increase?
Weighting and Aggregation is crucial (+15-20 points)Profit rate (+7-8 ppts) 6= Markup (+30-40 points)
2. ReallocationWeighting Matters: Input Weight
1.2
1.3
1.4
1.5
1.6
1960 1970 1980 1990 2000 2010
Sales Total cost
• See Grassi (2016) and Edmond, Midrigan and Xu (2019)
2. Reallocation
∆µt =∑i
mi,t−1∆µit︸ ︷︷ ︸∆ within
+∑i
µi,t−1∆mi,t︸ ︷︷ ︸∆ market share
+∑i
∆µi,t∆mi,t︸ ︷︷ ︸∆ cross-term
+∑
i∈Entry
µi,tmi,t −∑i∈Exit
µi,t−1mi,t−1︸ ︷︷ ︸net entry
1.20
1.30
1.40
1.50
1980 1990 2000 2010
Markup (benchmark)WithinReallocationNet Entry
See also Superstar Firms (Autor, Dorn, Katz, Patterson, Van Reenen (2018))
2. Reallocation
∆µt =∑i
mi,t−1∆µit︸ ︷︷ ︸∆ within
+∑i
µi,t−1∆mi,t︸ ︷︷ ︸∆ market share
+∑i
∆µi,t∆mi,t︸ ︷︷ ︸∆ cross-term
+∑
i∈Entry
µi,tmi,t −∑i∈Exit
µi,t−1mi,t−1︸ ︷︷ ︸net entry
1.20
1.30
1.40
1.50
1980 1990 2000 2010
Markup (benchmark)WithinReallocationNet Entry
See also Superstar Firms (Autor, Dorn, Katz, Patterson, Van Reenen (2018))
Facts
1. Heterogeneity: sharp rise for few firms; no rise for most
2. Reallocation of sales from low to high markup firms (2/3)
Technology Matters:
Fixed cost (SG&A) ↑Returns to Scale ↑
Magnitude of the Increase?
Weighting and Aggregation is crucial (+15-20 points)Profit rate (+7-8 ppts) 6= Markup (+30-40 points)
3. Technology MattersRise in Overhead (SG&A)
.12
.14
.16
.18
.2
.22
1960 1970 1980 1990 2000 2010
Facts
1. Heterogeneity: sharp rise for few firms; no rise for most
2. Reallocation of sales from low to high markup firms (2/3)
3. Technology Matters: Overhead cost (SG&A) ↑Magnitude of the Increase?
Weighting and Aggregation is crucial (+15-20 points)Profit rate (+7-8 ppts) 6= Markup (+30-40 points)
4. Magnitude of Increasea. Aggregation: Industry Averages: +20 points
1.1
1.2
1.3
1.4
1.5
1.6
1960 1970 1980 1990 2000 2010
Benchmark Aggregate MarkupAverages (By Industry, time-varying thetas)Averages (By Industry)Averages (Economy-wide)
• See also Hall (1988 and 2018)
4. Magnitude of Increaseb. Profit Rate: +7-8 ppt
0
.02
.04
.06
.08
.1
1960 1970 1980 1990 2000 2010
• Profits/Value Added: +15%
4. Magnitude of IncreaseProfit Rate vs Markup
• The profit rate:
πi =PiQi − C (Qi )
PiQi= 1− 1
µi
ACiMCi
⇒ With µ = 1.6 in 2016, implied profit rate is π = 1− 11.61 = 0.38!!
• This logic uses:1. Representative Firm Economy: but Aggregation (Jensen’s Inequality)2. Unchanged economies of scale (AC = MC ): but ACMC ↑ (Overhead ↑)
4. Magnitude of IncreaseProfit Rate vs Markup
• The profit rate:
πi =PiQi − C (Qi )
PiQi= 1− 1
µi
ACiMCi
⇒ With µ = 1.6 in 2016, implied profit rate is π = 1− 11.61 = 0.38!!
• This logic uses:1. Representative Firm Economy: but Aggregation (Jensen’s Inequality)2. Unchanged economies of scale (AC = MC ): but ACMC ↑ (Overhead ↑)
4. Magnitude of IncreaseProfit Rate vs Markup
0
.1
.2
.3
.4
1980 1990 2000 2010
Avg, No FCAggr, FCProfit Rate
Facts
1. Heterogeneity: sharp rise for few firms; no rise for most
2. Reallocation of sales from low to high markup firms (2/3)
3. Technology Matters: Overhead cost (SG&A) ↑4. Magnitude of the Increase?
a Weighting and Aggregation is crucialb Profit rate (+7-8 ppts) 6= Markup (+30-40 points)
∴ Only publicly traded firms (40% of GDP)
Facts
1. Heterogeneity: sharp rise for few firms; no rise for most
2. Reallocation of sales from low to high markup firms (2/3)
3. Technology Matters: Overhead cost (SG&A) ↑4. Magnitude of the Increase?
a Weighting and Aggregation is crucialb Profit rate (+7-8 ppts) 6= Markup (+30-40 points)
∴ Only publicly traded firms (40% of GDP)
Robustness: US CensusesManufacturing
1.5
1.6
1.7
1.8
1.9
1972 1977 1982 1987 1992 1997 2002 2007 2012
1.5
2
2.5
3
1972 1977 1982 1987 1992 1997 2002 2007 2012
Mw90Mw75Mw50
Global Markup134 countries; 70,000 firms; 1980-2016
1.1
1.2
1.3
1.4
1.5
1.6
1980 1990 2000 2010
GLOBAL
Markup Continents
1
1.2
1.4
1.6
1.8
1
1.2
1.4
1.6
1.8
1980 1990 2000 2010 1980 1990 2000 2010 1980 1990 2000 2010
EUROPE NORTH AMERICA SOUTH AMERICA
ASIA OCEANIA AFRICA
Germany
1
1.1
1.2
1.3
1.4
1980 1990 2000 2010
GERMANY
Quantifying Market Power:
Causes and Consequences
Quantifying Market Power
Market Power in General Equilibrium• Causes: need both
1. Market Structure: ABInBev→ labor reallocation down
2. Technology: Amazon Paradox→ fixed cost and productivity dispersion ⇒ markup dispersion
⇒ Net effect: Welfare loss
• Consequences: Secular Trends in Macro1. Wage Stagnation: equilibrium effect (not monopsony)2. Labor Share decline: at firm level3. Decline in Business Dynamism: incomplete passthrough4. Reallocation of sales towards high markup, large superstar firms
Quantifying Market Power
Market Power in General Equilibrium• Causes: need both
1. Market Structure: ABInBev→ labor reallocation down
2. Technology: Amazon Paradox→ fixed cost and productivity dispersion ⇒ markup dispersion
⇒ Net effect: Welfare loss• Consequences: Secular Trends in Macro
1. Wage Stagnation: equilibrium effect (not monopsony)2. Labor Share decline: at firm level3. Decline in Business Dynamism: incomplete passthrough4. Reallocation of sales towards high markup, large superstar firms
The Rise of Market Power
Jan Eeckhout
UPF Barcelona and UCL
BCCP ConferenceJune 21, 2019