globalization and firms: the challenge for theory · globalization and firms: the challenge for...
TRANSCRIPT
Globalization and Firms:The Challenge for Theory
J. Peter Neary
University of Oxford and CEPR
Chaire Theorie Economique et Organisation SocialeCollege de France, Paris
March 6, 2013
Peter Neary (Oxford) Globalization and Firms March 6, 2013 1 / 39
Introduction
Outline of the Talk
1 Introduction
2 Functional Form
3 Monopolistic Competition versus Oligopoly
4 Free Entry
5 General Equilibrium
6 Superstar Firms
7 Conclusion
Peter Neary (Oxford) Globalization and Firms March 6, 2013 2 / 39
Introduction
Motivation
Growing empirical evidence: large firms matter for trade
1st wave of micro data (1995-): Exporting firms are exceptional:
Larger, more productive
2nd wave: Even within exporters, large firms dominate:
Distribution of exporters is bimodalThe firms that matter (for most questions) are different: larger,multi-product, multi-destination
[Bernard et al. (JEP 2007), Mayer and Ottaviano (2007)]
Peter Neary (Oxford) Globalization and Firms March 6, 2013 3 / 39
Introduction
Motivation
Growing empirical evidence: large firms matter for trade
1st wave of micro data (1995-): Exporting firms are exceptional:
Larger, more productive
2nd wave: Even within exporters, large firms dominate:
Distribution of exporters is bimodalThe firms that matter (for most questions) are different: larger,multi-product, multi-destination
[Bernard et al. (JEP 2007), Mayer and Ottaviano (2007)]
Peter Neary (Oxford) Globalization and Firms March 6, 2013 3 / 39
Introduction
Motivation
Growing empirical evidence: large firms matter for trade
1st wave of micro data (1995-): Exporting firms are exceptional:
Larger, more productive
2nd wave: Even within exporters, large firms dominate:
Distribution of exporters is bimodalThe firms that matter (for most questions) are different: larger,multi-product, multi-destination
[Bernard et al. (JEP 2007), Mayer and Ottaviano (2007)]
Peter Neary (Oxford) Globalization and Firms March 6, 2013 3 / 39
Introduction
U.S. Evidence
Bernard et al. (JEP 2007):• Data on U.S. exporting firms 2000• By # of products & export destinationsBy # of products & export destinations
8Peter Neary (Oxford) Globalization and Firms March 6, 2013 4 / 39
Introduction
U.S. Evidence
Bernard et al. (JEP 2007):• Data on U.S. exporting firms 2000• By # of products & export destinations
5+ products:
By # of products & export destinations
25 9% f fi25.9% of firms
7Peter Neary (Oxford) Globalization and Firms March 6, 2013 5 / 39
Introduction
U.S. Evidence
Bernard et al. (JEP 2007):• Data on U.S. exporting firms 2000• By # of products & export destinations
5+ products:
By # of products & export destinations
25 9% f fi25.9% of firms
98.0% of export value
6Peter Neary (Oxford) Globalization and Firms March 6, 2013 6 / 39
Introduction
U.S. Evidence
Bernard et al. (JEP 2007):• Data on U.S. exporting firms 2000• By # of products & export destinations
5+ products:
By # of products & export destinations
25 9% f fi25.9% of firms
98.0% of export value
83.3% of employment
5Peter Neary (Oxford) Globalization and Firms March 6, 2013 7 / 39
Introduction
U.S. Evidence
Bernard et al. (JEP 2007):• Data on U.S. exporting firms 2000• By # of products & export destinations
5+ products:
By # of products & export destinations
25 9% f fi& 5+ dests.:
25.9% of firms
11.9% of firms
98.0% of export value
92.2% of export value
83.3% of employment
68.8% of employment
4Peter Neary (Oxford) Globalization and Firms March 6, 2013 8 / 39
Introduction
U.S. Evidence
Bernard et al. (JEP 2007):• Data on U.S. exporting firms 2000• By # of products & export destinations1 product
5+ products:
By # of products & export destinations
25 9% f fi
40.4% of firms
1 product, 1 destination only:
& 5+ dests.:25.9% of firms
11.9% of firms
98.0% of export value
0.20% of export value
7.0% of employment
92.2% of export value
83.3% of employment
68.8% of employment
4Peter Neary (Oxford) Globalization and Firms March 6, 2013 9 / 39
Introduction
Similarly in France
what I call the ‘second wave’ of micro data on firms and trade. The first wave,
from the mid-1990s onwards, showed that exporting firms are exceptional,
larger and more productive than average. More recent datasets in the second
wave have provided more disaggregated information on the activities of firms,
and have highlighted the degree of heterogeneity even within exporters.
Table 1 summarises some aspects of the distribution of manufacturing
exports for the US and France, adapted from Bernard et al. (2007) and Mayer
and Ottaviano (2007). Looking at the breakdowns by number of products sold
and number of foreign markets served, two features stand out. First is that
the distribution of firms is bimodal, with 40.4 per cent of US firms (29.6 per
cent of French firms) exporting only one product to only one market, while
11.9 per cent (23.3 per cent) export five or more products to five or more
markets. Second, the latter firms account for by far the bulk of the value of
exports, 92.2 per cent (87.3 per cent), so the distribution of export sales is
highly concentrated in the top exporters. If we ignore the number of destina-
tions and simply focus on the firms that export five or more products, we
find that they account for 25.9 per cent of US firms (34.3 per cent of French
firms) but an overwhelming 98.0 per cent (90.8 per cent) of exports. Bearing
in mind that non-exporting firms are excluded, these data suggest that the
largest exporting firms are different in kind from the majority. I am not the
first to draw attention to these features of the data. In particular, the work of
Xavier Gabaix (2005) on ‘granularity’, recently extended to international trade
by di Giovanni and Levchenko (2009), has also highlighted the importance of
large firms for aggregate behaviour. However, in their formal modelling those
authors stick with the assumption of monopolistic competition between firms.
They allow for large firms in one sense by assuming that the distribution of
firm productivities is Pareto with a high value for the dispersion parameter,
but nevertheless they continue to assume that all firms are infinitesimal in
scale. I want to go further: to try and put the grains into granularity.
TABLE 1Distribution of Manufacturing Exports by Number of Products and Markets
Number of US 2000 France 2003
Products Markets
% Share ofExportingFirms
% Share ofValue ofExports
% Share ofExportingFirms
% Shareof Valueof Exports
1 1 40.4 0.2 29.6 0.75+ 5+ 11.9 92.2 23.3 87.35+ 1+ 25.9 98.0 34.3 90.8
Notes:Data are extracted from Bernard et al. (2007, Table 4), and Mayer and Ottaviano (2007, Table A1). Productsare defined as 10-digit Harmonised System categories.
2009 The AuthorJournal compilation Blackwell Publishing Ltd. 2009
TWO AND A HALF THEORIES OF TRADE 3
Peter Neary (Oxford) Globalization and Firms March 6, 2013 10 / 39
Introduction
Motivation
Large firms important in other ways too:
They grow just as quickly as small ones
Conventional view that small firms grow faster . . .. . . suffers from a statistical illusion
[Berthou-Vicard (2013)]
They are older
They do more R&D
Peter Neary (Oxford) Globalization and Firms March 6, 2013 11 / 39
Introduction
Motivation
Large firms important in other ways too:
They grow just as quickly as small ones
Conventional view that small firms grow faster . . .. . . suffers from a statistical illusion
[Berthou-Vicard (2013)]
They are older
They do more R&D
Peter Neary (Oxford) Globalization and Firms March 6, 2013 11 / 39
Introduction
Motivation
Large firms important in other ways too:
They grow just as quickly as small ones
Conventional view that small firms grow faster . . .. . . suffers from a statistical illusion
[Berthou-Vicard (2013)]
They are older
They do more R&D
Peter Neary (Oxford) Globalization and Firms March 6, 2013 11 / 39
Introduction
Motivation
Large firms important in other ways too:
They grow just as quickly as small ones
Conventional view that small firms grow faster . . .. . . suffers from a statistical illusion
[Berthou-Vicard (2013)]
They are older
They do more R&D
Peter Neary (Oxford) Globalization and Firms March 6, 2013 11 / 39
Introduction
Motivation
Large firms important in other ways too:
They grow just as quickly as small ones
Conventional view that small firms grow faster . . .. . . suffers from a statistical illusion
[Berthou-Vicard (2013)]
They are older
They do more R&D
Peter Neary (Oxford) Globalization and Firms March 6, 2013 11 / 39
Introduction
So much for facts, what about theory?!
Mainstream model of firms in international trade:[Krugman (1980)-Melitz (2003)]
Strong assumptions about functional form
Market structure is monopolistic competition
. . . embedded in general equilibrium
Assumes rapid entry and exit
So: No “superstar” firms
Peter Neary (Oxford) Globalization and Firms March 6, 2013 12 / 39
Introduction
So much for facts, what about theory?!
Mainstream model of firms in international trade:[Krugman (1980)-Melitz (2003)]
Strong assumptions about functional form
Market structure is monopolistic competition
. . . embedded in general equilibrium
Assumes rapid entry and exit
So: No “superstar” firms
Peter Neary (Oxford) Globalization and Firms March 6, 2013 12 / 39
Introduction
So much for facts, what about theory?!
Mainstream model of firms in international trade:[Krugman (1980)-Melitz (2003)]
Strong assumptions about functional form
Market structure is monopolistic competition
. . . embedded in general equilibrium
Assumes rapid entry and exit
So: No “superstar” firms
Peter Neary (Oxford) Globalization and Firms March 6, 2013 12 / 39
Introduction
So much for facts, what about theory?!
Mainstream model of firms in international trade:[Krugman (1980)-Melitz (2003)]
Strong assumptions about functional form
Market structure is monopolistic competition
. . . embedded in general equilibrium
Assumes rapid entry and exit
So: No “superstar” firms
Peter Neary (Oxford) Globalization and Firms March 6, 2013 12 / 39
Introduction
So much for facts, what about theory?!
Mainstream model of firms in international trade:[Krugman (1980)-Melitz (2003)]
Strong assumptions about functional form
Market structure is monopolistic competition
. . . embedded in general equilibrium
Assumes rapid entry and exit
So: No “superstar” firms
Peter Neary (Oxford) Globalization and Firms March 6, 2013 12 / 39
Introduction
So much for facts, what about theory?!
Mainstream model of firms in international trade:[Krugman (1980)-Melitz (2003)]
Strong assumptions about functional form
Market structure is monopolistic competition
. . . embedded in general equilibrium
Assumes rapid entry and exit
So: No “superstar” firms
Peter Neary (Oxford) Globalization and Firms March 6, 2013 12 / 39
Introduction
Outline of the Talk
1 Introduction
2 Functional Form
3 Monopolistic Competition versus Oligopoly
4 Free Entry
5 General Equilibrium
6 Superstar Firms
7 Conclusion
Peter Neary (Oxford) Globalization and Firms March 6, 2013 13 / 39
Functional Form
Outline of the Talk
1 Introduction
2 Functional FormFrom General Demands to CESA Firm’s-Eye View of DemandCES and Super-ConvexityThe Demand ManifoldThe Pollak Demand FamilyGlobalization and Welfare with Pollak Preferences
3 Monopolistic Competition versus Oligopoly
4 Free Entry
5 General Equilibrium
6 Superstar Firms
7 Conclusion
Peter Neary (Oxford) Globalization and Firms March 6, 2013 14 / 39
Functional Form From General Demands to CES
From General Demands to CES
How to specify demands in monopolistic competition?
In principle: No restrictions [Chamberlin (1933)]
Key feature: Firms take not price but demand function as givenBut: Hard to get results or extend to general equilibrium
Breakthrough came with a specific tractable form: CES[Dixit-Stiglitz (1977)]
U =
[∫i∈Ω
ux(i)di]1/θ
, ux(i) = x(i)θ, 0 < θ < 1 (1)
⇔ x(i) = α[λp(i)]−1
1−θ (2)
Partial and general equilibrium linked cleanly by λEasy to work with theoretically, especially with symmetric goodsEasy to work with empirically: iso-elastic demand functionsBUT: Very special ...
Peter Neary (Oxford) Globalization and Firms March 6, 2013 15 / 39
Functional Form From General Demands to CES
From General Demands to CES
How to specify demands in monopolistic competition?
In principle: No restrictions [Chamberlin (1933)]
Key feature: Firms take not price but demand function as givenBut: Hard to get results or extend to general equilibrium
Breakthrough came with a specific tractable form: CES[Dixit-Stiglitz (1977)]
U =
[∫i∈Ω
ux(i)di]1/θ
, ux(i) = x(i)θ, 0 < θ < 1 (1)
⇔ x(i) = α[λp(i)]−1
1−θ (2)
Partial and general equilibrium linked cleanly by λEasy to work with theoretically, especially with symmetric goodsEasy to work with empirically: iso-elastic demand functionsBUT: Very special ...
Peter Neary (Oxford) Globalization and Firms March 6, 2013 15 / 39
Functional Form From General Demands to CES
From General Demands to CES
How to specify demands in monopolistic competition?
In principle: No restrictions [Chamberlin (1933)]
Key feature: Firms take not price but demand function as givenBut: Hard to get results or extend to general equilibrium
Breakthrough came with a specific tractable form: CES[Dixit-Stiglitz (1977)]
U =
[∫i∈Ω
ux(i)di]1/θ
, ux(i) = x(i)θ, 0 < θ < 1 (1)
⇔ x(i) = α[λp(i)]−1
1−θ (2)
Partial and general equilibrium linked cleanly by λEasy to work with theoretically, especially with symmetric goodsEasy to work with empirically: iso-elastic demand functionsBUT: Very special ...
Peter Neary (Oxford) Globalization and Firms March 6, 2013 15 / 39
Functional Form From General Demands to CES
From General Demands to CES
How to specify demands in monopolistic competition?
In principle: No restrictions [Chamberlin (1933)]
Key feature: Firms take not price but demand function as givenBut: Hard to get results or extend to general equilibrium
Breakthrough came with a specific tractable form: CES[Dixit-Stiglitz (1977)]
U =
[∫i∈Ω
ux(i)di]1/θ
, ux(i) = x(i)θ, 0 < θ < 1 (1)
⇔ x(i) = α[λp(i)]−1
1−θ (2)
Partial and general equilibrium linked cleanly by λEasy to work with theoretically, especially with symmetric goodsEasy to work with empirically: iso-elastic demand functionsBUT: Very special ...
Peter Neary (Oxford) Globalization and Firms March 6, 2013 15 / 39
Functional Form From General Demands to CES
From General Demands to CES
How to specify demands in monopolistic competition?
In principle: No restrictions [Chamberlin (1933)]
Key feature: Firms take not price but demand function as givenBut: Hard to get results or extend to general equilibrium
Breakthrough came with a specific tractable form: CES[Dixit-Stiglitz (1977)]
U =
[∫i∈Ω
ux(i)di]1/θ
, ux(i) = x(i)θ, 0 < θ < 1 (1)
⇔ x(i) = α[λp(i)]−1
1−θ (2)
Partial and general equilibrium linked cleanly by λEasy to work with theoretically, especially with symmetric goodsEasy to work with empirically: iso-elastic demand functions
BUT: Very special ...
Peter Neary (Oxford) Globalization and Firms March 6, 2013 15 / 39
Functional Form From General Demands to CES
From General Demands to CES
How to specify demands in monopolistic competition?
In principle: No restrictions [Chamberlin (1933)]
Key feature: Firms take not price but demand function as givenBut: Hard to get results or extend to general equilibrium
Breakthrough came with a specific tractable form: CES[Dixit-Stiglitz (1977)]
U =
[∫i∈Ω
ux(i)di]1/θ
, ux(i) = x(i)θ, 0 < θ < 1 (1)
⇔ x(i) = α[λp(i)]−1
1−θ (2)
Partial and general equilibrium linked cleanly by λEasy to work with theoretically, especially with symmetric goodsEasy to work with empirically: iso-elastic demand functionsBUT: Very special ...
Peter Neary (Oxford) Globalization and Firms March 6, 2013 15 / 39
Functional Form A Firm’s-Eye View of Demand
A Firm’s-Eye View of Demand
Perceived inverse demand function:p = p(x) p′ < 0
Firm cares about:
1 Slope/Elasticity:
ε(x) ≡ − p(x)xp′(x) > 0
2 Curvature/Convexity:
ρ(x) ≡ −xp′′(x)p′(x)
4
4
3
2
11
0-2 -1 0 1 2 3
Alternative measures of slope and curvature . . .Skip
Peter Neary (Oxford) Globalization and Firms March 6, 2013 16 / 39
Functional Form A Firm’s-Eye View of Demand
A Firm’s-Eye View of Demand
Perceived inverse demand function:p = p(x) p′ < 0
Firm cares about:
1 Slope/Elasticity:
ε(x) ≡ − p(x)xp′(x) > 0
2 Curvature/Convexity:
ρ(x) ≡ −xp′′(x)p′(x)
4
4
3
2
11
0-2 -1 0 1 2 3
Alternative measures of slope and curvature . . .Skip
Peter Neary (Oxford) Globalization and Firms March 6, 2013 16 / 39
Functional Form A Firm’s-Eye View of Demand
A Firm’s-Eye View of Demand
Perceived inverse demand function:p = p(x) p′ < 0
Firm cares about:
1 Slope/Elasticity:
ε(x) ≡ − p(x)xp′(x) > 0
2 Curvature/Convexity:
ρ(x) ≡ −xp′′(x)p′(x)
4
4
3
2
11
0-2 -1 0 1 2 3
Alternative measures of slope and curvature . . .Skip
Peter Neary (Oxford) Globalization and Firms March 6, 2013 16 / 39
Functional Form A Firm’s-Eye View of Demand
A Firm’s-Eye View of Demand
Perceived inverse demand function:p = p(x) p′ < 0
Firm cares about:
1 Slope/Elasticity:
ε(x) ≡ − p(x)xp′(x) > 0
2 Curvature/Convexity:
ρ(x) ≡ −xp′′(x)p′(x)
4
4
3
2
11
0-2 -1 0 1 2 3
Alternative measures of slope and curvature . . .Skip
Peter Neary (Oxford) Globalization and Firms March 6, 2013 16 / 39
Functional Form A Firm’s-Eye View of Demand
The Admissible Region
For a monopoly firm:
First-order condition:p+ xp′ = c ≥ 0 ⇒ ε ≥ 1
Second-order condition:2p′ + xp′′ < 0 ⇒ ρ < 2
Both less stringent in oligopolyDetails
Peter Neary (Oxford) Globalization and Firms March 6, 2013 17 / 39
Functional Form A Firm’s-Eye View of Demand
The Admissible Region
For a monopoly firm:
First-order condition:p+ xp′ = c ≥ 0 ⇒ ε ≥ 1
Second-order condition:2p′ + xp′′ < 0 ⇒ ρ < 2
Both less stringent in oligopolyDetails
4.0
4.0
3.0
2.0
1 01.0
0.0-2.0 -1.0 0.0 1.0 2.0 3.0
Peter Neary (Oxford) Globalization and Firms March 6, 2013 17 / 39
Functional Form A Firm’s-Eye View of Demand
The Admissible Region
For a monopoly firm:
First-order condition:p+ xp′ = c ≥ 0 ⇒ ε ≥ 1
Second-order condition:2p′ + xp′′ < 0 ⇒ ρ < 2
Both less stringent in oligopolyDetails
4.0
4.0
3.0
2.0
1 01.0
0.0-2.0 -1.0 0.0 1.0 2.0 3.0
Peter Neary (Oxford) Globalization and Firms March 6, 2013 17 / 39
Functional Form CES and Super-Convexity
CES Demands
In general, both ε and ρ vary withsales
Exception: CES/iso-elastic case:
p = βx−1/σ
⇒ ε = σ, ρ = σ+1σ > 1
⇒ ε = 1ρ−1
4.0
CES4.0
3.0
2.0Cobb-Douglas
1 0
Cobb Douglas
1.0
0.0-2.0 -1.0 0.0 1.0 2.0 3.0
Cobb-Douglas: ε = 1, ρ = 2; just on boundary of both FOC and SOC
Peter Neary (Oxford) Globalization and Firms March 6, 2013 18 / 39
Functional Form CES and Super-Convexity
CES Demands
In general, both ε and ρ vary withsales
Exception: CES/iso-elastic case:
p = βx−1/σ
⇒ ε = σ, ρ = σ+1σ > 1
⇒ ε = 1ρ−1
4.0
CES4.0
3.0
2.0Cobb-Douglas
1 0
Cobb Douglas
1.0
0.0-2.0 -1.0 0.0 1.0 2.0 3.0
Cobb-Douglas: ε = 1, ρ = 2; just on boundary of both FOC and SOC
Peter Neary (Oxford) Globalization and Firms March 6, 2013 18 / 39
Functional Form CES and Super-Convexity
CES Demands
In general, both ε and ρ vary withsales
Exception: CES/iso-elastic case:
p = βx−1/σ
⇒ ε = σ, ρ = σ+1σ > 1
⇒ ε = 1ρ−1
4.0
CES4.0
3.0
2.0Cobb-Douglas
1 0
Cobb Douglas
1.0
0.0-2.0 -1.0 0.0 1.0 2.0 3.0
Cobb-Douglas: ε = 1, ρ = 2; just on boundary of both FOC and SOC
Peter Neary (Oxford) Globalization and Firms March 6, 2013 18 / 39
Functional Form CES and Super-Convexity
Super-Convexity
[Mrazova-Neary (2011)]
Definition :
p(x) is superconvex IFF log[p(x)] isconvex in log(x)
⇔ p(x) more convex than a CESdemand function with the sameelasticity
SCSuper-
4.0 Convex
A B
3.0 SubConvex
A B
2.0
C
1.0
0.0-2.0 -1.0 0.0 1.0 2.0 3.0
Peter Neary (Oxford) Globalization and Firms March 6, 2013 19 / 39
Functional Form CES and Super-Convexity
Super-Convexity
[Mrazova-Neary (2011)]
Definition :
p(x) is superconvex IFF log[p(x)] isconvex in log(x)
⇔ p(x) more convex than a CESdemand function with the sameelasticity
SCSuper-
4.0 Convex
A B
3.0 SubConvex
A B
2.0
C
1.0
0.0-2.0 -1.0 0.0 1.0 2.0 3.0
Peter Neary (Oxford) Globalization and Firms March 6, 2013 19 / 39
Functional Form CES and Super-Convexity
Super-Convexity
[Mrazova-Neary (2011)]
Definition :
p(x) is superconvex IFF log[p(x)] isconvex in log(x)
⇔ p(x) more convex than a CESdemand function with the sameelasticity
SCSuper-
4.0 Convex
A B
3.0 SubConvex
A B
2.0
C
1.0
0.0-2.0 -1.0 0.0 1.0 2.0 3.0
Peter Neary (Oxford) Globalization and Firms March 6, 2013 19 / 39
Functional Form CES and Super-Convexity
Super-Convexity and Sales
p(x) superconvex:⇔ ε increasing in sales: εx ≥ 0.
εx = εx
[ρ− ε+1
ε
]= ε
x
[ρ− ρCES
]1 ε decreases with sales to left2 ε is independent of sales only
along CES/SC locus3 ε increases with sales to right
Which is most plausible?
εx < 0: “Marshall’s 2nd Law of Demand”!
Marshall (1920), Krugman (1979)Linear/Quadratic, LES/Stone-Geary, CARA, etc.
The comparative-statics analogue of a phase diagram:
Arrows indicate direction as sales rise
Peter Neary (Oxford) Globalization and Firms March 6, 2013 20 / 39
Functional Form CES and Super-Convexity
Super-Convexity and Sales
p(x) superconvex:⇔ ε increasing in sales: εx ≥ 0.
εx = εx
[ρ− ε+1
ε
]= ε
x
[ρ− ρCES
]
1 ε decreases with sales to left2 ε is independent of sales only
along CES/SC locus3 ε increases with sales to right
Which is most plausible?
εx < 0: “Marshall’s 2nd Law of Demand”!
Marshall (1920), Krugman (1979)Linear/Quadratic, LES/Stone-Geary, CARA, etc.
The comparative-statics analogue of a phase diagram:
Arrows indicate direction as sales rise
Peter Neary (Oxford) Globalization and Firms March 6, 2013 20 / 39
Functional Form CES and Super-Convexity
Super-Convexity and Sales
p(x) superconvex:⇔ ε increasing in sales: εx ≥ 0.
εx = εx
[ρ− ε+1
ε
]= ε
x
[ρ− ρCES
]1 ε decreases with sales to left2 ε is independent of sales only
along CES/SC locus3 ε increases with sales to right
4.0
SC
4.0
3.0
2.0
1 01.0
0.0-2.0 -1.0 0.0 1.0 2.0 3.0
Which is most plausible?
εx < 0: “Marshall’s 2nd Law of Demand”!
Marshall (1920), Krugman (1979)Linear/Quadratic, LES/Stone-Geary, CARA, etc.
The comparative-statics analogue of a phase diagram:
Arrows indicate direction as sales rise
Peter Neary (Oxford) Globalization and Firms March 6, 2013 20 / 39
Functional Form CES and Super-Convexity
Super-Convexity and Sales
p(x) superconvex:⇔ ε increasing in sales: εx ≥ 0.
εx = εx
[ρ− ε+1
ε
]= ε
x
[ρ− ρCES
]1 ε decreases with sales to left2 ε is independent of sales only
along CES/SC locus3 ε increases with sales to right
4.0
SC
4.0
3.0
2.0
1 01.0
0.0-2.0 -1.0 0.0 1.0 2.0 3.0 Which is most plausible?
εx < 0: “Marshall’s 2nd Law of Demand”!
Marshall (1920), Krugman (1979)Linear/Quadratic, LES/Stone-Geary, CARA, etc.
The comparative-statics analogue of a phase diagram:
Arrows indicate direction as sales rise
Peter Neary (Oxford) Globalization and Firms March 6, 2013 20 / 39
Functional Form CES and Super-Convexity
Super-Convexity and Sales
p(x) superconvex:⇔ ε increasing in sales: εx ≥ 0.
εx = εx
[ρ− ε+1
ε
]= ε
x
[ρ− ρCES
]1 ε decreases with sales to left2 ε is independent of sales only
along CES/SC locus3 ε increases with sales to right
4.0
SC
4.0
3.0
2.0
1 01.0
0.0-2.0 -1.0 0.0 1.0 2.0 3.0 Which is most plausible?
εx < 0: “Marshall’s 2nd Law of Demand”!
Marshall (1920), Krugman (1979)Linear/Quadratic, LES/Stone-Geary, CARA, etc.
The comparative-statics analogue of a phase diagram:
Arrows indicate direction as sales rise
Peter Neary (Oxford) Globalization and Firms March 6, 2013 20 / 39
Functional Form CES and Super-Convexity
Super-Convexity and Sales
p(x) superconvex:⇔ ε increasing in sales: εx ≥ 0.
εx = εx
[ρ− ε+1
ε
]= ε
x
[ρ− ρCES
]1 ε decreases with sales to left2 ε is independent of sales only
along CES/SC locus3 ε increases with sales to right
4.0
SC
4.0
3.0
2.0
1 01.0
0.0-2.0 -1.0 0.0 1.0 2.0 3.0 Which is most plausible?
εx < 0: “Marshall’s 2nd Law of Demand”!
Marshall (1920), Krugman (1979)Linear/Quadratic, LES/Stone-Geary, CARA, etc.
The comparative-statics analogue of a phase diagram:
Arrows indicate direction as sales risePeter Neary (Oxford) Globalization and Firms March 6, 2013 20 / 39
Functional Form The Demand Manifold
The Demand Manifold
For most demand functions:
ε(x) and ρ(x) can be solvedfor ε = E(ρ) ≡ ε [x(ρ)]
The “Demand Manifold”
Special cases:
CES: Collapses to a pointLinear: Collapses to a line
4.0
CES4.0
3.0
2.0Cobb-Douglas
1 0
Cobb Douglas
1.0
0.0-2.0 -1.0 0.0 1.0 2.0 3.0
When is the Demand Manifold invariant to shocks?
p = p(x, φ) ⇒ ε = ε(x, φ), ρ = ρ(x, φ) ⇒ E(ρ, φ) = ε [X(ρ, φ), φ]
E is independent of φ in CES and linear cases. Does this generalize?
Peter Neary (Oxford) Globalization and Firms March 6, 2013 21 / 39
Functional Form The Demand Manifold
The Demand Manifold
For most demand functions:
ε(x) and ρ(x) can be solvedfor ε = E(ρ) ≡ ε [x(ρ)]The “Demand Manifold”
Special cases:
CES: Collapses to a pointLinear: Collapses to a line
4.0
CES4.0
3.0
2.0Cobb-Douglas
1 0
Cobb Douglas
1.0
0.0-2.0 -1.0 0.0 1.0 2.0 3.0
When is the Demand Manifold invariant to shocks?
p = p(x, φ) ⇒ ε = ε(x, φ), ρ = ρ(x, φ) ⇒ E(ρ, φ) = ε [X(ρ, φ), φ]
E is independent of φ in CES and linear cases. Does this generalize?
Peter Neary (Oxford) Globalization and Firms March 6, 2013 21 / 39
Functional Form The Demand Manifold
The Demand Manifold
For most demand functions:
ε(x) and ρ(x) can be solvedfor ε = E(ρ) ≡ ε [x(ρ)]The “Demand Manifold”
Special cases:
CES: Collapses to a pointLinear: Collapses to a line
4.0
CES4.0
3.0
2.0Cobb-Douglas
1 0
Cobb Douglas
1.0
0.0-2.0 -1.0 0.0 1.0 2.0 3.0
When is the Demand Manifold invariant to shocks?
p = p(x, φ) ⇒ ε = ε(x, φ), ρ = ρ(x, φ) ⇒ E(ρ, φ) = ε [X(ρ, φ), φ]
E is independent of φ in CES and linear cases. Does this generalize?
Peter Neary (Oxford) Globalization and Firms March 6, 2013 21 / 39
Functional Form The Demand Manifold
The Demand Manifold
For most demand functions:
ε(x) and ρ(x) can be solvedfor ε = E(ρ) ≡ ε [x(ρ)]The “Demand Manifold”
Special cases:
CES: Collapses to a pointLinear: Collapses to a line
4.0
CES4.0
3.0
2.0Cobb-Douglas
1 0
Cobb Douglas
1.0
0.0-2.0 -1.0 0.0 1.0 2.0 3.0
When is the Demand Manifold invariant to shocks?
p = p(x, φ) ⇒ ε = ε(x, φ), ρ = ρ(x, φ) ⇒ E(ρ, φ) = ε [X(ρ, φ), φ]
E is independent of φ in CES and linear cases. Does this generalize?
Peter Neary (Oxford) Globalization and Firms March 6, 2013 21 / 39
Functional Form The Demand Manifold
The Demand Manifold
For most demand functions:
ε(x) and ρ(x) can be solvedfor ε = E(ρ) ≡ ε [x(ρ)]The “Demand Manifold”
Special cases:
CES: Collapses to a pointLinear: Collapses to a line
4.0
CES4.0
3.0
2.0Cobb-Douglas
1 0
Cobb Douglas
1.0
0.0-2.0 -1.0 0.0 1.0 2.0 3.0
When is the Demand Manifold invariant to shocks?
p = p(x, φ) ⇒ ε = ε(x, φ), ρ = ρ(x, φ) ⇒ E(ρ, φ) = ε [X(ρ, φ), φ]
E is independent of φ in CES and linear cases. Does this generalize?
Peter Neary (Oxford) Globalization and Firms March 6, 2013 21 / 39
Functional Form The Demand Manifold
The Demand Manifold
For most demand functions:
ε(x) and ρ(x) can be solvedfor ε = E(ρ) ≡ ε [x(ρ)]The “Demand Manifold”
Special cases:
CES: Collapses to a pointLinear: Collapses to a line
4.0
CES4.0
3.0
2.0Cobb-Douglas
1 0
Cobb Douglas
1.0
0.0-2.0 -1.0 0.0 1.0 2.0 3.0
When is the Demand Manifold invariant to shocks?
p = p(x, φ) ⇒ ε = ε(x, φ), ρ = ρ(x, φ) ⇒ E(ρ, φ) = ε [X(ρ, φ), φ]
E is independent of φ in CES and linear cases. Does this generalize?
Peter Neary (Oxford) Globalization and Firms March 6, 2013 21 / 39
Functional Form The Demand Manifold
The Demand Manifold
For most demand functions:
ε(x) and ρ(x) can be solvedfor ε = E(ρ) ≡ ε [x(ρ)]The “Demand Manifold”
Special cases:
CES: Collapses to a pointLinear: Collapses to a line
4.0
CES4.0
3.0
2.0Cobb-Douglas
1 0
Cobb Douglas
1.0
0.0-2.0 -1.0 0.0 1.0 2.0 3.0
When is the Demand Manifold invariant to shocks?
p = p(x, φ) ⇒ ε = ε(x, φ), ρ = ρ(x, φ) ⇒ E(ρ, φ) = ε [X(ρ, φ), φ]
E is independent of φ in CES and linear cases. Does this generalize?
Peter Neary (Oxford) Globalization and Firms March 6, 2013 21 / 39
Functional Form The Pollak Demand Family
The Pollak Demand Family
x = γ + αp1
θ−1 , (x− γ)(1− θ) > 0
ε = 2−θ1−θ
1ρ
θ → −∞: CARA
θ ∈ (−∞, 1): “Translated CES”:
θ ∈ (−∞, 0): TCES-I
θ = 0: Stone-Geary LES
θ ∈ (0, 1): TCES-II
θ ∈ (1,∞): “GeneralizedQuadratic”:
θ ∈ (1, 2): Sub-Quadraticθ = 2: Quadraticθ ∈ (2,∞): Super-Quadratic
Peter Neary (Oxford) Globalization and Firms March 6, 2013 22 / 39
Functional Form The Pollak Demand Family
The Pollak Demand Family
x = γ + αp1
θ−1 , (x− γ)(1− θ) > 0
ε = 2−θ1−θ
1ρ
θ → −∞: CARA
θ ∈ (−∞, 1): “Translated CES”:
θ ∈ (−∞, 0): TCES-I
θ = 0: Stone-Geary LES
θ ∈ (0, 1): TCES-II
θ ∈ (1,∞): “GeneralizedQuadratic”:
θ ∈ (1, 2): Sub-Quadraticθ = 2: Quadraticθ ∈ (2,∞): Super-Quadratic
4.0
SM
4.0
3.0SC
2.0
1 01.0
SPT
= ∞
0.0-2.00 -1.00 0.00 1.00 2.00 3.00
SPT
Peter Neary (Oxford) Globalization and Firms March 6, 2013 22 / 39
Functional Form The Pollak Demand Family
The Pollak Demand Family
x = γ + αp1
θ−1 , (x− γ)(1− θ) > 0
ε = 2−θ1−θ
1ρ
θ → −∞: CARA
θ ∈ (−∞, 1): “Translated CES”:
θ ∈ (−∞, 0): TCES-I
θ = 0: Stone-Geary LES
θ ∈ (0, 1): TCES-II
θ ∈ (1,∞): “GeneralizedQuadratic”:
θ ∈ (1, 2): Sub-Quadraticθ = 2: Quadraticθ ∈ (2,∞): Super-Quadratic
4.0
SM
4.0
3.0SC
2.0
1 01.0
SPT
= ∞= 3
0.0-2.00 -1.00 0.00 1.00 2.00 3.00
SPT
Peter Neary (Oxford) Globalization and Firms March 6, 2013 22 / 39
Functional Form The Pollak Demand Family
The Pollak Demand Family
x = γ + αp1
θ−1 , (x− γ)(1− θ) > 0
ε = 2−θ1−θ
1ρ
θ → −∞: CARA
θ ∈ (−∞, 1): “Translated CES”:
θ ∈ (−∞, 0): TCES-I
θ = 0: Stone-Geary LES
θ ∈ (0, 1): TCES-II
θ ∈ (1,∞): “GeneralizedQuadratic”:
θ ∈ (1, 2): Sub-Quadraticθ = 2: Quadraticθ ∈ (2,∞): Super-Quadratic
4.0
SM
4.0
3.0SC
2.0
1 01.0
SPT
= ∞= 0= 3
0.0-2.00 -1.00 0.00 1.00 2.00 3.00
SPT
Peter Neary (Oxford) Globalization and Firms March 6, 2013 22 / 39
Functional Form The Pollak Demand Family
The Pollak Demand Family
x = γ + αp1
θ−1 , (x− γ)(1− θ) > 0
ε = 2−θ1−θ
1ρ
θ → −∞: CARA
θ ∈ (−∞, 1): “Translated CES”:
θ ∈ (−∞, 0): TCES-I
θ = 0: Stone-Geary LES
θ ∈ (0, 1): TCES-II
θ ∈ (1,∞): “GeneralizedQuadratic”:
θ ∈ (1, 2): Sub-Quadraticθ = 2: Quadraticθ ∈ (2,∞): Super-Quadratic
4.0
SM
4.0
3.0SC
2.0= 0.8
1 0= 0.5
1.0
SPT
= ∞= 0= 3
0.0-2.00 -1.00 0.00 1.00 2.00 3.00
SPT
Peter Neary (Oxford) Globalization and Firms March 6, 2013 22 / 39
Functional Form The Pollak Demand Family
The Pollak Demand Family
x = γ + αp1
θ−1 , (x− γ)(1− θ) > 0
ε = 2−θ1−θ
1ρ
θ → −∞: CARA
θ ∈ (−∞, 1): “Translated CES”:
θ ∈ (−∞, 0): TCES-I
θ = 0: Stone-Geary LES
θ ∈ (0, 1): TCES-II
θ ∈ (1,∞): “GeneralizedQuadratic”:
θ ∈ (1, 2): Sub-Quadratic
θ = 2: Quadraticθ ∈ (2,∞): Super-Quadratic
4.0
SM
4.0
3.0SC
2.0= 0.8
1 0
= 1.33
= 0.51.0
SPT
= 1.5 = ∞= 0= 3
0.0-2.00 -1.00 0.00 1.00 2.00 3.00
SPT
Peter Neary (Oxford) Globalization and Firms March 6, 2013 22 / 39
Functional Form The Pollak Demand Family
The Pollak Demand Family
x = γ + αp1
θ−1 , (x− γ)(1− θ) > 0
ε = 2−θ1−θ
1ρ
θ → −∞: CARA
θ ∈ (−∞, 1): “Translated CES”:
θ ∈ (−∞, 0): TCES-I
θ = 0: Stone-Geary LES
θ ∈ (0, 1): TCES-II
θ ∈ (1,∞): “GeneralizedQuadratic”:
θ ∈ (1, 2): Sub-Quadraticθ = 2: Quadratic
θ ∈ (2,∞): Super-Quadratic
4.0
SM
4.0
3.0SC
2.0= 0.8
1 0
= 1.33
= 0.51.0
SPT
= 1.5 = 2 = ∞= 0= 3
0.0-2.00 -1.00 0.00 1.00 2.00 3.00
SPT
Peter Neary (Oxford) Globalization and Firms March 6, 2013 22 / 39
Functional Form The Pollak Demand Family
The Pollak Demand Family
x = γ + αp1
θ−1 , (x− γ)(1− θ) > 0
ε = 2−θ1−θ
1ρ
θ → −∞: CARA
θ ∈ (−∞, 1): “Translated CES”:
θ ∈ (−∞, 0): TCES-I
θ = 0: Stone-Geary LES
θ ∈ (0, 1): TCES-II
θ ∈ (1,∞): “GeneralizedQuadratic”:
θ ∈ (1, 2): Sub-Quadraticθ = 2: Quadraticθ ∈ (2,∞): Super-Quadratic
4.0
SM
4.0
3.0SC
2.0= 0.8
1 0
= 1.33
= 0.51.0
SPT
= 1.5 = 2 = 3 = ∞= 0= 3
0.0-2.00 -1.00 0.00 1.00 2.00 3.00
SPT
Peter Neary (Oxford) Globalization and Firms March 6, 2013 22 / 39
Functional Form The Pollak Demand Family
The Pollak Demand Family
x = γ + αp1
θ−1 , (x− γ)(1− θ) > 0
ε = 2−θ1−θ
1ρ
θ → −∞: CARA
θ ∈ (−∞, 1): “Translated CES”:
θ ∈ (−∞, 0): TCES-I
θ = 0: Stone-Geary LES
θ ∈ (0, 1): TCES-II
θ ∈ (1,∞): “GeneralizedQuadratic”:
θ ∈ (1, 2): Sub-Quadraticθ = 2: Quadraticθ ∈ (2,∞): Super-Quadratic
4.0
SM
4.0
3.0SC
2.0= 0.8
1 0
= 1.33
= 0.51.0
SPT
= 1.5 = 2 = 3 = ∞= 0= 3
0.0-2.00 -1.00 0.00 1.00 2.00 3.00
SPT
Peter Neary (Oxford) Globalization and Firms March 6, 2013 23 / 39
Functional Form Globalization and Welfare with Pollak Preferences
Globalization and Welfare with Pollak Preferences
Monopolistic competition
General equilibriumPollak Preferences
Gains from globalization:[Rise in number of countries k]
U =
[1− (ε− 1)2
ε2 (2− ρ)
]k
= 1/ε in CES case
Sufficient condition for U > 0:ρ < ε+1
ε i.e., subconvexity.
Welfare can fall if preferences are sufficiently superconvex
Diversity rises a lot, but prices increase
Welfare rises by more for lower ε and ρ
Peter Neary (Oxford) Globalization and Firms March 6, 2013 24 / 39
Functional Form Globalization and Welfare with Pollak Preferences
Globalization and Welfare with Pollak Preferences
Monopolistic competition
General equilibriumPollak Preferences
Gains from globalization:[Rise in number of countries k]
U =
[1− (ε− 1)2
ε2 (2− ρ)
]k
= 1/ε in CES case
Sufficient condition for U > 0:ρ < ε+1
ε i.e., subconvexity.
Welfare can fall if preferences are sufficiently superconvex
Diversity rises a lot, but prices increase
Welfare rises by more for lower ε and ρ
Peter Neary (Oxford) Globalization and Firms March 6, 2013 24 / 39
Functional Form Globalization and Welfare with Pollak Preferences
Globalization and Welfare with Pollak Preferences
Monopolistic competition
General equilibriumPollak Preferences
Gains from globalization:[Rise in number of countries k]
U =
[1− (ε− 1)2
ε2 (2− ρ)
]k
= 1/ε in CES case
Sufficient condition for U > 0:ρ < ε+1
ε i.e., subconvexity.
Welfare can fall if preferences are sufficiently superconvex
Diversity rises a lot, but prices increase
Welfare rises by more for lower ε and ρ
Peter Neary (Oxford) Globalization and Firms March 6, 2013 24 / 39
Functional Form Globalization and Welfare with Pollak Preferences
Globalization and Welfare with Pollak Preferences
Monopolistic competition
General equilibriumPollak Preferences
Gains from globalization:[Rise in number of countries k]
U =
[1− (ε− 1)2
ε2 (2− ρ)
]k
= 1/ε in CES case
Sufficient condition for U > 0:ρ < ε+1
ε i.e., subconvexity.
4.0
4.00ˆ U
3.0
2.0
1 01.0
0.0-2.00 -1.00 0.00 1.00 2.00 3.00
Welfare can fall if preferences are sufficiently superconvex
Diversity rises a lot, but prices increase
Welfare rises by more for lower ε and ρ
Peter Neary (Oxford) Globalization and Firms March 6, 2013 24 / 39
Functional Form Globalization and Welfare with Pollak Preferences
Globalization and Welfare with Pollak Preferences
Monopolistic competition
General equilibriumPollak Preferences
Gains from globalization:[Rise in number of countries k]
U =
[1− (ε− 1)2
ε2 (2− ρ)
]k
= 1/ε in CES case
Sufficient condition for U > 0:ρ < ε+1
ε i.e., subconvexity.
4.0
4.00ˆ U
3.0
2.0
1 01.0
0.0-2.00 -1.00 0.00 1.00 2.00 3.00
Welfare can fall if preferences are sufficiently superconvex
Diversity rises a lot, but prices increase
Welfare rises by more for lower ε and ρPeter Neary (Oxford) Globalization and Firms March 6, 2013 24 / 39
Functional Form Globalization and Welfare with Pollak Preferences
Globalization and Welfare with Pollak Preferences
Monopolistic competition
General equilibriumPollak Preferences
Gains from globalization:[Rise in number of countries k]
U =
[1− (ε− 1)2
ε2 (2− ρ)
]k
= 1/ε in CES case
Sufficient condition for U > 0:ρ < ε+1
ε i.e., subconvexity.
4.0
4.00ˆ UkU ˆ5.0ˆ
3.0 kU ˆ8.0ˆ
2.0
1 01.0
0.0-2.00 -1.00 0.00 1.00 2.00 3.00
Welfare can fall if preferences are sufficiently superconvex
Diversity rises a lot, but prices increase
Welfare rises by more for lower ε and ρPeter Neary (Oxford) Globalization and Firms March 6, 2013 24 / 39
Monopolistic Competition versus Oligopoly
Outline of the Talk
1 Introduction
2 Functional Form
3 Monopolistic Competition versus Oligopoly
4 Free Entry
5 General Equilibrium
6 Superstar Firms
7 Conclusion
Peter Neary (Oxford) Globalization and Firms March 6, 2013 25 / 39
Monopolistic Competition versus Oligopoly
Monopolistic Competition
“New” trade theory borrowed from half of IO only
IO (Industrial Organization): Partial equilibrium onlyTrade: Oligopoly squeezed out by monopolistic competition
Monopolistic competition more plausible than perfect competition . . .
Differentiated productsIncreasing returnsSo: successful in explaining intra-industry trade
. . . but not much!
Firms are infinitesimalNo strategic behaviour
Peter Neary (Oxford) Globalization and Firms March 6, 2013 26 / 39
Monopolistic Competition versus Oligopoly
Monopolistic Competition
“New” trade theory borrowed from half of IO only
IO (Industrial Organization): Partial equilibrium onlyTrade: Oligopoly squeezed out by monopolistic competition
Monopolistic competition more plausible than perfect competition . . .
Differentiated productsIncreasing returnsSo: successful in explaining intra-industry trade
. . . but not much!
Firms are infinitesimalNo strategic behaviour
Peter Neary (Oxford) Globalization and Firms March 6, 2013 26 / 39
Monopolistic Competition versus Oligopoly
Monopolistic Competition
“New” trade theory borrowed from half of IO only
IO (Industrial Organization): Partial equilibrium onlyTrade: Oligopoly squeezed out by monopolistic competition
Monopolistic competition more plausible than perfect competition . . .
Differentiated productsIncreasing returnsSo: successful in explaining intra-industry trade
. . . but not much!
Firms are infinitesimalNo strategic behaviour
Peter Neary (Oxford) Globalization and Firms March 6, 2013 26 / 39
Free Entry
Outline of the Talk
1 Introduction
2 Functional Form
3 Monopolistic Competition versus Oligopoly
4 Free Entry
5 General Equilibrium
6 Superstar Firms
7 Conclusion
Peter Neary (Oxford) Globalization and Firms March 6, 2013 27 / 39
Free Entry
Free Entry
Standard trade models assume instantaneous entry and exit
Entry and exit are much less important in the short runFrench firms adjusted along intensive margin in the crisis
[Bricongne, Fontagne, Gaulier and Taglioni (JIE 2012)]
U.S. firms adjust more along extensive margin the longer the timehorizon
[Bernard et al. (2007)]
Entry and exit are much less important for large firmsMelitz model assumes that probability of “death” is independent offirm size or productivityBut: very successful firms are typically older
Entry and exit are much less important for value of exports than forthe number of firms
Even with free entry, “natural oligopoly” may prevail if fixed costs canbe chosen endogenously[Dasgupta-Stiglitz (EJ 1980), Gabszewicz-Thisse (JET 1980), Shaked-Sutton (Em
1983)]
Peter Neary (Oxford) Globalization and Firms March 6, 2013 28 / 39
Free Entry
Free Entry
Standard trade models assume instantaneous entry and exitEntry and exit are much less important in the short run
French firms adjusted along intensive margin in the crisis[Bricongne, Fontagne, Gaulier and Taglioni (JIE 2012)]
U.S. firms adjust more along extensive margin the longer the timehorizon
[Bernard et al. (2007)]
Entry and exit are much less important for large firmsMelitz model assumes that probability of “death” is independent offirm size or productivityBut: very successful firms are typically older
Entry and exit are much less important for value of exports than forthe number of firms
Even with free entry, “natural oligopoly” may prevail if fixed costs canbe chosen endogenously[Dasgupta-Stiglitz (EJ 1980), Gabszewicz-Thisse (JET 1980), Shaked-Sutton (Em
1983)]
Peter Neary (Oxford) Globalization and Firms March 6, 2013 28 / 39
Free Entry
Free Entry
Standard trade models assume instantaneous entry and exitEntry and exit are much less important in the short run
French firms adjusted along intensive margin in the crisis[Bricongne, Fontagne, Gaulier and Taglioni (JIE 2012)]
U.S. firms adjust more along extensive margin the longer the timehorizon
[Bernard et al. (2007)]
Entry and exit are much less important for large firmsMelitz model assumes that probability of “death” is independent offirm size or productivityBut: very successful firms are typically older
Entry and exit are much less important for value of exports than forthe number of firms
Even with free entry, “natural oligopoly” may prevail if fixed costs canbe chosen endogenously[Dasgupta-Stiglitz (EJ 1980), Gabszewicz-Thisse (JET 1980), Shaked-Sutton (Em
1983)]
Peter Neary (Oxford) Globalization and Firms March 6, 2013 28 / 39
Free Entry
Free Entry
Standard trade models assume instantaneous entry and exitEntry and exit are much less important in the short run
French firms adjusted along intensive margin in the crisis[Bricongne, Fontagne, Gaulier and Taglioni (JIE 2012)]
U.S. firms adjust more along extensive margin the longer the timehorizon
[Bernard et al. (2007)]
Entry and exit are much less important for large firmsMelitz model assumes that probability of “death” is independent offirm size or productivityBut: very successful firms are typically older
Entry and exit are much less important for value of exports than forthe number of firms
Even with free entry, “natural oligopoly” may prevail if fixed costs canbe chosen endogenously[Dasgupta-Stiglitz (EJ 1980), Gabszewicz-Thisse (JET 1980), Shaked-Sutton (Em
1983)]
Peter Neary (Oxford) Globalization and Firms March 6, 2013 28 / 39
Free Entry
Free-Entry Cournot: Market Size and Firm Numbers
1
0.00
1.00
2.00
3.00
4.00
5.00
6.00
7.00
8.00
9.00
10.00
0.0 1.0 2.0 3.0 4.0 5.0 6.0
n
s
Cournot Competition: Equilibrium n as a Function of Market Size
Real
Peter Neary (Oxford) Globalization and Firms March 6, 2013 29 / 39
Free Entry
Free-Entry Cournot with Integer Firms
1
0.00
1.00
2.00
3.00
4.00
5.00
6.00
7.00
8.00
9.00
10.00
0.0 1.0 2.0 3.0 4.0 5.0 6.0
n
s
Cournot Competition: Equilibrium n as a Function of Market Size
Real
Integer
Peter Neary (Oxford) Globalization and Firms March 6, 2013 30 / 39
Free Entry
Natural Oligopoly: Market Size and Firm Numbers
1
0.00
1.00
2.00
3.00
4.00
5.00
6.00
7.00
8.00
9.00
10.00
0.0 1.0 2.0 3.0 4.0 5.0 6.0
n
s
Equilibrium Real n as a Function of Market Size
h=0.0
h=0.2
h=0.4
h=0.6
h=0.8
h=1.0
h=1.2
h=1.4
h=1.6
h=1.8
h=2.0
Peter Neary (Oxford) Globalization and Firms March 6, 2013 31 / 39
Free Entry
Natural Oligopoly with Integer Firms
1
0
1
2
3
4
5
6
7
8
9
10
0.0 1.0 2.0 3.0 4.0 5.0 6.0
n
s
Equilibrium Integer n as a Function of Market Size
h=0.0
h=0.2
h=0.4
h=0.6
h=0.8
h=1.0
h=1.2
h=1.4
h=1.6
h=1.8
h=2.0
Peter Neary (Oxford) Globalization and Firms March 6, 2013 32 / 39
General Equilibrium
Outline of the Talk
1 Introduction
2 Functional Form
3 Monopolistic Competition versus Oligopoly
4 Free Entry
5 General Equilibrium
6 Superstar Firms
7 Conclusion
Peter Neary (Oxford) Globalization and Firms March 6, 2013 33 / 39
General Equilibrium
General Equilibrium
Core questions in trade are general equilibrium
In the sense of requiring interactions between goods and factor markets
Theoretical barriers to putting “OLigopoly” into “GE”: “GOLE”
Do large firms affect wages? national income? the price level?
Resolution: View firms as “large in the small, small in the large”[Hart (QJE, 1982), Neary (JEEA 2003)]
Like monopolistic competition but more firms in each sector
U =
∫i∈Ω
ux(i)di ⇔ x(i) = x[λp(i)] (3)
Application: Cross-border mergers [Neary (REStud 2007)]
Mergers may be for strategic or synergistic reasonsIn partial equilibrium, strategic mergers must lower consumer surplusIn GE, they can raise welfare if resources are reallocated to moreefficient firms
Peter Neary (Oxford) Globalization and Firms March 6, 2013 34 / 39
General Equilibrium
General Equilibrium
Core questions in trade are general equilibrium
In the sense of requiring interactions between goods and factor markets
Theoretical barriers to putting “OLigopoly” into “GE”: “GOLE”
Do large firms affect wages? national income? the price level?
Resolution: View firms as “large in the small, small in the large”[Hart (QJE, 1982), Neary (JEEA 2003)]
Like monopolistic competition but more firms in each sector
U =
∫i∈Ω
ux(i)di ⇔ x(i) = x[λp(i)] (3)
Application: Cross-border mergers [Neary (REStud 2007)]
Mergers may be for strategic or synergistic reasonsIn partial equilibrium, strategic mergers must lower consumer surplusIn GE, they can raise welfare if resources are reallocated to moreefficient firms
Peter Neary (Oxford) Globalization and Firms March 6, 2013 34 / 39
General Equilibrium
General Equilibrium
Core questions in trade are general equilibrium
In the sense of requiring interactions between goods and factor markets
Theoretical barriers to putting “OLigopoly” into “GE”: “GOLE”
Do large firms affect wages? national income? the price level?
Resolution: View firms as “large in the small, small in the large”[Hart (QJE, 1982), Neary (JEEA 2003)]
Like monopolistic competition but more firms in each sector
U =
∫i∈Ω
ux(i)di ⇔ x(i) = x[λp(i)] (3)
Application: Cross-border mergers [Neary (REStud 2007)]
Mergers may be for strategic or synergistic reasonsIn partial equilibrium, strategic mergers must lower consumer surplusIn GE, they can raise welfare if resources are reallocated to moreefficient firms
Peter Neary (Oxford) Globalization and Firms March 6, 2013 34 / 39
General Equilibrium
General Equilibrium
Core questions in trade are general equilibrium
In the sense of requiring interactions between goods and factor markets
Theoretical barriers to putting “OLigopoly” into “GE”: “GOLE”
Do large firms affect wages? national income? the price level?
Resolution: View firms as “large in the small, small in the large”[Hart (QJE, 1982), Neary (JEEA 2003)]
Like monopolistic competition but more firms in each sector
U =
∫i∈Ω
ux(i)di ⇔ x(i) = x[λp(i)] (3)
Application: Cross-border mergers [Neary (REStud 2007)]
Mergers may be for strategic or synergistic reasonsIn partial equilibrium, strategic mergers must lower consumer surplusIn GE, they can raise welfare if resources are reallocated to moreefficient firms
Peter Neary (Oxford) Globalization and Firms March 6, 2013 34 / 39
Superstar Firms
Outline of the Talk
1 Introduction
2 Functional Form
3 Monopolistic Competition versus Oligopoly
4 Free Entry
5 General Equilibrium
6 Superstar Firms
7 Conclusion
Peter Neary (Oxford) Globalization and Firms March 6, 2013 35 / 39
Superstar Firms
Superstar Firms
Evidence suggests large firms are different in more than just scale
Bimodality in the data suggests a modelling strategy:
Oligopoly of multi-product firm . . .. . . plus a monopolistically competitive fringe
Technically: Each large firm produces a finite measure of goodsAll products are differentiated and of measure zeroFits with recent work on multi-product firms in trade
[Eckel and Neary (REStud 2010), Bernard et al. (QJE 2011)]
Some progress to date:
“David and Goliath”: Neary (WE 2009), Shimomura and Thisse (RJE
2012), Parenti (2012)
Peter Neary (Oxford) Globalization and Firms March 6, 2013 36 / 39
Superstar Firms
Superstar Firms
Evidence suggests large firms are different in more than just scale
Bimodality in the data suggests a modelling strategy:
Oligopoly of multi-product firm . . .. . . plus a monopolistically competitive fringe
Technically: Each large firm produces a finite measure of goodsAll products are differentiated and of measure zeroFits with recent work on multi-product firms in trade
[Eckel and Neary (REStud 2010), Bernard et al. (QJE 2011)]
Some progress to date:
“David and Goliath”: Neary (WE 2009), Shimomura and Thisse (RJE
2012), Parenti (2012)
Peter Neary (Oxford) Globalization and Firms March 6, 2013 36 / 39
Conclusion
Outline of the Talk
1 Introduction
2 Functional Form
3 Monopolistic Competition versus Oligopoly
4 Free Entry
5 General Equilibrium
6 Superstar Firms
7 Conclusion
Peter Neary (Oxford) Globalization and Firms March 6, 2013 37 / 39
Conclusion
The Best Model for a Globalized World?
Not one but many
Plausible, falsifiable, simple (but not too much so!)
Some desirable features:
Not too reliant on special functional formsRecognise strategic behaviour by large firmsAllow for general equilibrium. . . and for free entry, at least by small firmsAllow for superstar firms
Peter Neary (Oxford) Globalization and Firms March 6, 2013 38 / 39
Conclusion
The Best Model for a Globalized World?
Not one but many
Plausible, falsifiable, simple (but not too much so!)
Some desirable features:
Not too reliant on special functional formsRecognise strategic behaviour by large firmsAllow for general equilibrium. . . and for free entry, at least by small firmsAllow for superstar firms
Peter Neary (Oxford) Globalization and Firms March 6, 2013 38 / 39
Conclusion
The Best Model for a Globalized World?
Not one but many
Plausible, falsifiable, simple (but not too much so!)
Some desirable features:
Not too reliant on special functional formsRecognise strategic behaviour by large firmsAllow for general equilibrium. . . and for free entry, at least by small firmsAllow for superstar firms
Peter Neary (Oxford) Globalization and Firms March 6, 2013 38 / 39
Conclusion
Thanks and Acknowledgements
Thank you for listening. Comments welcome!
The research leading to these results has received funding from the European Research Council under the European Union’s
Seventh Framework Programme (FP7/2007-2013), ERC grant agreement no. 295669. The contents reflect only the authors’
views and not the views of the ERC or the European Commission, and the European Union is not liable for any use that may be
made of the information contained therein.
Peter Neary (Oxford) Globalization and Firms March 6, 2013 39 / 39