Innovation and Competition: Theory, Evidence and Policy for the Great Recession

1
Innovation and Competition Theory, Evidence and
Policy for the Great Recession

• Federico Etro
• University of Milan, Bicocca
• Dynamic Competition Lecture
• Osaka, November 27 2009

2
Analysis of Business creation RD in micro- and
macro- phenomena- Many issues under a common
perspective- Tool of analysis the Endogenous
Market Structures Approach

• Theoretical principles
• Empirical evidence
• Policy applications

3

• Endogenous market structures (EMSs)
• strategic interactions
• and endogenous entry
• (not free entry with perfect competition, but
  endogenous number of firms due to rational entry
  in imperfectly competitive markets)
• wider research on EMSs in
• industrial organization (Sutton, 1991, 1998,
  etc),
• macroeconomics, trade, innovation growth
• industrial and trade policy, macroeconomic
  policy,
• contract theory and corporate finance
• Look at two books..

4
Advertising!
5
Traditional approach to market analysis
Performance
Conduct
Structure
6
EMS approach to market analysis
7
Example to keep in mind

• Isoelastic utility
• degree of substitutability
• Population of size L
• can derive direct (and inverse) demand Di
• assume fixed cost F and marginal cost c
• Gross profits

8
Example to keep in mind

• Monopolistic competition à la Dixit-Stiglitz
• Optimal mark up
• Number of firms
• Under homogenous goods with zero fixed costs
  marginal cost pricing and indeterminacy of the
  market structure

9
Cournot competition (homogenous goods)

• Equilibrium mark up
• Endogenous entry

10
With dynamic entry/exit (Etro-Colciago, 2010,
Economic Journal)

• Entry and profits are procyclical mark ups are
  countercyclical
• A Competition Effect helps the propagation of
  shocks (beyond the neoclassical mechanism) boom
  gt entry gt lower markups gt higher wages gt C and L
  go up
• Steady state EMSs depend on
• Market size, L (and productivity A in general
  equilibrium)
• Entry cost, F
• Substitutability,
• Bankruptcy rate,
• Discounting, r

11
US (HP) detrended data
12
VARanalysison US data
13
Implications for the crisis

• Important (supply-side) mechanism of propagation?
• Stock market crash gt Business destruction gt
  concentration gt higher mark ups and lower wages gt
  lower consumption and employment gt lower profits
  
• .. and trade collapse

14
Test of the market size effects

• In the monopolistic competition (Dixit-Stiglitz)
  approach
• In the EMSs approach (Cournot competition with
  homogenous goods)
• Joint work with Dirk Czarnitzki tests

15
Market size vs Number of firms (German NACE
3-digit)
16
Market size vs Number of firms (German NACE
4-digit)
17
Trade the Krugman model

• Integration between two equal countries (LL)
• Optimal mark up
• Number of firms
• Gains from trade gains from variety

18
Trade the Cournot model (with homogenous goods)

• Integration between two equal countries (LL)
• Optimal mark up
• Number of firms
• Gains from trade gains from competition (lower
  prices)
• See Ghironi and Melitz (2005, QJE) for a dynamic
  model

19

• Is this only about competition in the market?
• or EMSs have something to say about competition
  for the market?

20
EMSs and Innovation

• When competition is for the market
• Firms decide how much to invest taking into
  account expected profits and each others
• Endogenous entry determines
• Number of investors
• Individual investment
• Aggregate RD and rate of technological progress

21

• The simplest example of competition for the
  market
• z probability of innovation
• V value of (IPR protected) innovation
• Quadratic costs of investment
• Expected profits
• EMSs

22
what about patentholders?

• What is the role of market leaders in investing
  in RD and promoting technological progress?
• Commonly held view (based on Arrow, 1962)
• firms invest more in a competitive market where
  entry pressure is stronger
• incumbents tend to be less innovative than their
  followers
• incumbents do not invest when entry is free
• persistence of dominance is signal of market
  power and lack of entry pressure

23
Innovation by Leaders

• EMSs results
• RD spending per firm can decrease with entry and
  when entry is endogenous
• incumbents tend to be more innovative than their
  followers when entry is endogenous (Etro, 2004,
  Economic Journal)
• persistence of dominance is a signal of
  competition when there is entry pressure
• In the above example

24
Summing up....

• Two sufficient conditions where incumbent has
  incentive to invest in RD and invest more than
  others
• Leadership of incumbent monopolist
• Endogenous entry for outsiders in the innovation
  race
• Testable hypothesis
• Investment of incumbent leader is larger than
  investment of the average firm when entry is
  endogenous

25
Some evidence (from joint work with Czarnitzki
and Kraft)

• 2005 survey of the Mannheim Innovation Panel
• German Innovation Survey since 1992
• German part of the Community Innovation Survey
  (CIS)
• We focus on manufacturing sector
• 1,857 firm-level observations

26
Variables

• Dependent Variable
• RD intensity RD/Sales 100
• Right-hand side
• Incumbent (dummy) a firm that indicates that it
  is larger than its competitors in main product
  market (INC)
• ENTRY 4 categories
• low to high entry pressure in main product market
• Identify incumbents under entry threat
• INCENTRY

27
Other controls

• Firm size in terms of employment in t-1
• and its squared value
• Capital Intensity (in t-1) physical
  assets/employment
• 12 industry dummies
• Additionally patent stock (since 1978) as
  control
• higher protection of previous RD may lead to
  higher current investment
• allows us to use some retrospective data to
  control for unobserved heterogeneity

28
(No Transcript)
29
Econometrics

• We estimate censored regression models, as not
  all firms invest in RD
• homoscedastic and heteroscedastic models yield
  same conclusions
• We also test for feedback effects from current
  RD investment on perceived entry threat, as this
  would bias the estimates

30
Table 2 Heteroskedastic Tobit model on RDINT
31
Support for the EMSs results

• Entry threat reduces the RD investment of the
  average firm,
• but market leaders do invest more in RD than the
  average firm, the larger the entry threat is.
• Consequently, under these conditions, incumbents
  are more likely to innovate eventually.
• This may explain the persistence of the leadership

32
Implications for growth

• RD policy to subsidize RD always (Etro, 2008,
  Journal Macroeconomics)
• International coordination of subsidies to
  internalize positive spillovers across
  heterogenous countries (see Alesina, Angeloni
  Etro, 2005, American Economic Review)

33
Implications for the crisis

• Need to support RD Investment through
• RD subsidies and
• IPRs protection
• Different antitrust attitude toward high-tech
  leaders

34
Normative analysis I

• Suboptimality of EMSs and Dynamic inefficiency
• Countercyclical fiscal policy and tax rates
  (implications for the crisis supply
  side-intervention when demand side doesnt work
  see Japanese experience)
• Monetary policy aimed at minimizing the impact of
  price frictions on incentives to invest in RD
  and business creation (see Bilbiie, Ghironi
  Melitz, 2007, NBERMa)

35
Normative analysis II

• Trade policy export subsidization always optimal
  under endogenous entry (Etro, 2010, International
  Economic Review)
• Optimal export subsidy
• inverse of demand elasticity (opposite of
  Lerner optimal export tax)
• lower in case of imperfect susbtitutabiliy
• prohibitive in case of constant or decreasing
  marginal costs
• Is also the equilibrium subsidy when multiple
  countries adopt it
• Implications Active protectionism may be good!
• Exchange rate policy to promote exports
• Another case for IPRs protection and RD
  subsidies!

36
EMSs and Competition Policy

• post-Chicago approach was mainly focused on
  games with an incumbent and an entrant or a fixed
  number of firms (exogenous market structures)
• Ex. predatory pricing, tying, vertical
  restraints, mergers
• endogenizing entry some results change or need a
  different interpretation
• on a general approach Etro (2006, RAND Journal
  of Economics 2008, Economic Journal)
• on mergers Davidson-Mukherjee (2007, IJIO),
  Erkal-Piccinin (2008)
• on RD investment and predatory pricing
  Kovac-Vinogradov-Zigic (2010, JEDC)
• on technology transfers Creane-Konishi (2009,
  IJIO)
• on tying and vertical restraints
• (timing of) entry becomes crucial in antitrust
  investigations

37
Tying and EMSs

• The Chicago approach (single monopoly profit
  theorem) associates tying with efficiency reasons
  
• The post-Chicago approach to tying starts with
  Whinston (1990, AER)
• tying must be anti-competitive when there are a
  monopolist in the primary market and a duopoly in
  the secondary market
• tying strengthens competition
• gt the only purpose of tying is entry deterrence

38
EMSs analysis of tying

• Tying to strengthen competition (reduce prices)
  is profitable
• when there is product differentiation in the
  secondary market
• when multiple secondary goods can be bought at
  the same time
• when entry in the secondary market is endogenous
• when demand for the primary good is close to
  demand for the bundle
• total welfare increases and consumer surplus is
  unchanged (with Dixit-Stiglitz demand)

39
A famous case to keep in mind Windows-IE

• The primary market (OSs) is led by Windows
• the secondary market (browsers) is characterized
  by
• product differentiation (IE, Firefox, Opera,
  Chrome,..)
• multihoming (multiple browsers can be tried and
  used at the same time)
• endogenous entry (dynamic competitive process
  with entry of new browsers and expansion of
  competitors)
• demand for Windows is close to the demand for
  the bundle Windowsbrowser (few want PCs without
  browser)

40
A famous case Windows-IESource Net
Applications Data
41
Conclusions on Antitrust

• Endogenous number of firms overturn some results
  of the post-Chicago approach
• Entry conditions are crucial to verify an
  abusive strategy
• For instance, tying is a normal competitive
  (price-reducing) equilibrium strategy when the
  secondary market is characterized by endogenous
  entry

42
Final remarks

• EMSs are a more realistic representation of real
  markets imperfectly competitive
• EMSs improve the performance of neoclassical
  models of business cycle and trade and growth
  explaining better the process of business
  creation and RD
• EMSs provide new implications for macroeconomic
  policy, trade policy and RD policy supply-side
  intervention
• EMSs support many results of the Chicago school
  entry regulates dominant firms

43
?????