Empirical IO

Introduction and Overview

Structure-Conduct-Performance
paradigm
Based on the lectures of
Dr Christos Genakos (University of Cambridge)

References and background readings:

Pepall, L. Richards, D., and Norman G (2005), Industrial

Organization: Contemporary Theory and Practice, Thomson.
Church J. and R. Ware (2000), Industrial Organization: A Strategic
Approach, McGrawHill.
Motta M (2004) Competition Policy: Theory and Practice, Cambridge
University Press.
Perloff, Karp and Golan, (2007) Estimating Market Power and
Strategies, Cambridge University Press.
Einav, Liran and Levin Jonathan (2010) Empirical Industrial
Organization.
A Progress Report, Journal of Economics

Perspectives, 2, 145-162.

OUTLINE
1. Introduction and a brief history of IO
2. The BIG Questions in IO
3. The Structure-Conduct-Performance paradigm
4. Structure-Conduct-Performance in practice
5. Structure-Conduct-Performance: results
6. Structure-Conduct-Performance: critique

Introduction to Industrial Organization (IO)

A definition of IO: "Industrial organization is concerned with
the workings of markets and industries, in particular the way
firms compete with each other.
Aim:
Develop skills to make more informed decisions and
judgments about issues relating to the industry
become more efficient in analysing industries by identifying
key issues.

Objectives:
To get an improved understanding of static and
dynamic problems faced by firms:
1. internally, organizing production within the firm- The Theory
of the Firm
2. externally, how firms compete in the marketplace - The
Theory of Markets

A Brief History of Industrial Organization

Harvard Tradition (1940-1960; Joe Bain)
Structure-Conduct-Performance paradigm
Regressions on a cross section of industries to identify
correlations (performance = f(concentration, barriers to
entry))
Argued that high concentration was bad for consumers, and

paved the way for much antitrust legislation

Main weaknesses: (i) Assumes that structure

(concentration) is exogenous. (ii) Assumes away important
differences between industries.

A Brief History of Industrial Organization

Chicago School (1960-1980)
More careful application of econometric techniques
Use different market structures to understand different
industries or markets
Markets work!!!
Monopoly
is much more often alleged than confirmed

When monopoly does exist, it is often transitory; entry (or

just the threat of entry) is important

A Brief History of Industrial Organization

Game Theory (1980-1990)
Emphasis on strategic decision making
Modelled mathematically using Nash equilibrium concept
Produces a huge proliferation of models are often very
intuitive theoretically
However,
it is difficult to know which model is the right one

for a real world industry

A Brief History of Industrial Organization

New Empirical I.O. (1990 - )

Integrates theory and econometrics into structural

models

More complex empirical models that are computationally

intensive

I.O economists have now a much richer toolbox to

analyse markets

OUTLINE
1. Introduction and a brief history of IO
2. The BIG Questions in IO
3. The Structure-Conduct-Performance paradigm
4. Structure-Conduct-Performance in practice
5. Structure-Conduct-Performance: results
6. Structure-Conduct-Performance: critique

Why do we care?
Industrial Organization is centred around four questions:
1. Is there market power?
2. How do firms acquire and maintain market power?

3. What
are the implications of market power?
4. Is there a role for public policy regarding market power?

OUTLINE
1. Introduction and a brief history of IO
2. The BIG Questions in IO
3. The Structure-Conduct-Performance paradigm
4. Structure-Conduct-Performance in practice
5. Structure-Conduct-Performance: results
6. Structure-Conduct-Performance: critique

Structure-Conduct-Performance framework
SCP paradigm: Stable causal relationship between the
structure of an industry, firm conduct and market
performance.
Typical SCP study consists of two steps:
1) Obtain a measure of performance through direct
measurement (rather than estimation) and several
measures of industry structure (concentration, barriers to
entry,
unionization etc).

2) Regress performance measures on the various structure

measures to explain the differences in market
performance across industries i
i=+1CONC+2 B.E.i1+3 B.E.i2 +++1 B.E.iN+i

Structure-Conduct-Performance framework
Two main assumptions to establish a statistically and
conceptually meaningful relationship between structure
and market power:
1) The various industry structure measures are exogenous,
i.e. structure affects performance but not the other way
around.
2) Implied
degree of symmetry in conduct holds across

industries,
i.e. changes in the structural variables must

have the same average effect on market power in all

markets.

OUTLINE
1. Introduction and a brief history of IO
2. The BIG Questions in IO
3. The Structure-Conduct-Performance paradigm
4. Structure-Conduct-Performance in practice
5. Structure-Conduct-Performance: results
6. Structure-Conduct-Performance: critique

Structure-Conduct-Performance in practice
Performance measures:
Rate of return

ROI, ROS, ROE

key concern is whether the reported RoR captures

economic or accounting profits

Price-cost margin or Lerner index=(p-mc)/p

data is hard to come by, so most use AVC

MC

Tobins q which is the ratio of the firms market value to the

replacement cost of its assets

Need accurate measures of market value and

replacement cost of capital, >1 greater implied profits

Structure-Conduct-Performance in practice
Market Concentration:
Assume there are n firms producing a homogeneous good,
same mc, total output Q q1 q2 ... qn
Profit for firm i is:

i p (Q ) qi mcqi

dp
mc MC
Firms compete a la Cournot: MR p qi
dQ

p mc
1
s
Solve the n-firm equilibrium and rewrite: L p n
Generalize to allow for different mci:
2

p mci
si
HHI
L si

i
i

Structure-Conduct-Performance in practice
Concentration measures:

Herfindahl-Hirschman Index (HHI):

HHI si

i 1

Varies between 0(=perfect competition) and 1(=monopoly)

Takes into consideration the absolute number and size
distribution of firms

Concentration
Ratios:

Most common: CR4, CR8

CRm si
i 1

Structure-Conduct-Performance in practice
Barriers to entry:

Minimum Efficiency Scale

Advertising

Capital sunk capital investments

R&D

Unionization:
unions may be able to extract higher wages

hence
reducing profits associated with mkt power

Buyer Power: just as seller concentration is important,

buyer concentration may lead to lower prices and less
mkt power for sellers

OUTLINE
1. Introduction and a brief history of IO
2. The BIG Questions in IO
3. The Structure-Conduct-Performance paradigm
4. Structure-Conduct-Performance in practice
5. Structure-Conduct-Performance: results
6. Structure-Conduct-Performance: critique

Structure-Conduct-Performance: results
Hypothesis 1: market power should increase as
concentration increases
Weiss (1974) reviews literature prior to 1970s, most studies found
a positive relationship, but the effect is small (10% increase in
C4 resulted in 1.21% increase in price-cost margins).
Schmalensee (1989) who surveys the literature after Weiss, cast
doubt on the sign and whether the effect is statistically
significant.

There
is some suggestion that the relationship between

profitability and concentration is discontinuous; critical level of

concentration 70% (Bain, 1951).

Firm-level data studies confirm that link weak if it exists at all and
its the presence of a large second or third firm that greatly
reduces price-cost margins (Kwoka and Ravenscraft, 1985).

Structure-Conduct-Performance: results
Hypothesis 2: the larger the barriers to entry, the greater the
exercise of market power
Effect of barriers to entry more robust and significant than
concentration.
MES, capital intensity, R&D to sales, advertising to sales, all
positively correlated with profits, though highly correlated with
each other.
Increased buyer concentration lowers price-cost margins

Unionism

has a significant negative effect on price-cost margins

Salinger (1984) suggested that link between concentration and
profits exists only if there are barriers to entry:

i=+CONCi [2 B.E.i1+3 B.E.i2 +++1 B.E.iN]+i

Test significance of interaction: none of the coef or sum significant

OUTLINE
1. Introduction and a brief history of IO
2. The BIG Questions in IO
3. The Structure-Conduct-Performance paradigm
4. Structure-Conduct-Performance in practice
5. Structure-Conduct-Performance: results
6. Structure-Conduct-Performance: critique

Structure-Conduct-Performance: Critique
Measurement Problems
RoR
1. Capital is not valued appropriately; historical cost vs.
replacement cost, book value vs. economic value
2. Depreciation is measured improperly; economic rental rate on
capital after depreciation, econ vs. accounting depreciation
3. Valuing advertising and R&D

4. Adjusting
for risk, debt vs. equity

5. Adjustments for inflation

6. Capitalized monopoly profits inappropriately included by using
book value
7. Pre-tax instead of after-tax RoR often calculated

Structure-Conduct-Performance: Critique
Measurement Problems
Price Cost Margins
Instead of MC, AVC is used:
PCM=(Sales Revenue-Payroll costs-Material costs)/Sales Revenue
BUT substituting AVC for MC may cause serious bias.

Structure-Conduct-Performance: Critique
Measurement Problems
Tobins q
1. Avoid the problems with estimating RoR or MC
2. We need meaningful measures of both market value and
replacement cost of firms assets
3. Firm value: equities+debt; issues with efficient market
hypothesis, timing of evaluation

4. Hard
to obtain estimate of replacement costs unless markets for

used equipment exist

5. Much harder to evaluate intangible assets like advertising, R&D

and human capital, usually ignore those and hence ratio>1

Structure-Conduct-Performance: Critique
Measurement Problems

Concentration Measures
1. How do we define a market? Economic vs. national statistic
agency definition.
2. Boundaries of an economic market should include all the firms
and their products that interact to determine prices, demandside (product) and supply-side (geographic) substitutes.

importance of cross-price elasticities!!!

3. The

4. Standard Industry Classification (or new North American ICS)

either too general or too specific (51 sector, 513 subsector,
5133 industry group, etc)
5. No import-export data

Structure-Conduct-Performance: Critique

Structure-Conduct-Performance: Critique
Conceptual Problems
1. Long run vs. Short run
Assumed stable relationship between mkt structure and long-run
profitability
2. Symmetric Industry Effects

Elasticity of industry demand not included, assumed the

same!!!!

3. What does a positive correlation between concentration and

profitability mean?

More mkt power in concentrated industries or more efficient firms?

Implication is that a firms success is explained better by its own
mkt share rather than just by industry concentration

Structure-Conduct-Performance: Critique
This critique is mostly associated with Demsetz (1973,
1974) and is commonly known as the market power
versus efficiency debate.
Demsetzs basic argument runs as follows. Industries
become concentrated because more efficient firms
grow at the expense of less efficient firms. A positive
correlation between concentration and profitability at

the industry level might simply be the result of a

positive
concentration-profitability relationship for large
(and efficient) firms and an insignificant concentrationprofitability relation for small (and inefficient) firms.

Structure-Conduct-Performance: Critique
4. Causality

The SCP is estimated econometrically usually with crosssection multiple regression, across industries
- Unidirectional causality
Profits = f(concentration, Barriers, X)
- Later,
other equations added to account for feed-back

For example:
Concentration = g(barriers, mkt size, Z)
Advertising-sales ratio = h(concentration, profits, barriers)
And then simultaneous systems TSLS, see Strickland &
Weiss for typical.

Structure-Conduct-Performance: References
*Schmalensee, R. (1989) Inter-industry studies of structure and
performance, Handbook of Industrial Organization, 951-1009.

provides a comprehensive survey and assessment of SCP studies

Geroski, P. (1981) Specification and Testing the Profits-Concentration
Relationship: Some experiments for the UK, Economica 48:279-88.

for UK evidence
*Salinger (1990) The concentration-Margins Relationship Reconsidered,
Brookings
Papers on Economic Activity on Microeconomics, 287-321.

for a careful, illustrative relative recent example.

Geroski (1988) In Pursuit of Monopoly Power: Recent Quantitative Work in
Industrial Economics, Journal of Applied Econometrics 3: 107-123.

for the potential value of the SCP approach for policy and antitrust
enforcement

Next time: New Empirical Industrial

Organization (NEIO) and Industry Models of
Market Power
*Bresnahan, T. (1982) The Oligopoly Solution is Identified,
Economic Letters, 10: 87-92.
*Bresnahan, T. (1989) Empirical Studies of Industries with Market
Power, Handbook of Industrial Organization, 1011-1057.
Corts, K. (1999) Conduct Parameters and the Measurement of
Market
Power, Journal of Econometrics, 88:227-250.

Genesove,
D. and Mullin, W. (1998) Testing Static Oligopoly
Models: Conduct and Cost in the Sugar Industry, 1890-1914,
Rand Journal of Economics, 29:355-377.