Você está na página 1de 99

Negotiating for the Market

Joshua Gans
NBER Summer Institute
July 2010
Creation without Destruction
Creation without Destruction

• Competition: Technological entrepreneurship can displace product market


incumbents
• Entrepeneurship implies creative destruction
Creation without Destruction

• Competition: Technological entrepreneurship can displace product market


incumbents
• Entrepeneurship implies creative destruction

• Cooperation: Technological entrepreneurship may reinforce existing market


power
• Entrepreneurship leads to creation but no destruction
Creation without Destruction

• Competition: Technological entrepreneurship can displace product market


incumbents
• Entrepeneurship implies creative destruction

• Cooperation: Technological entrepreneurship may reinforce existing market


power
• Entrepreneurship leads to creation but no destruction

Key insight: Competition vs cooperation is not a given but a choice ... the
result of negotiations between start-ups and incumbents.
Broad research agenda
Broad research agenda

• Builds on research program with Scott Stern (MIT) and David Hsu
(Wharton) ...
Broad research agenda

• Builds on research program with Scott Stern (MIT) and David Hsu
(Wharton) ...

• Goal: use insights from bargaining theory to understand the


commercialisation choices of entrepreneurial start-ups and the impact of
these on the rate of innovation
Broad research agenda

• Builds on research program with Scott Stern (MIT) and David Hsu
(Wharton) ...

• Goal: use insights from bargaining theory to understand the


commercialisation choices of entrepreneurial start-ups and the impact of
these on the rate of innovation

• This paper ...


• What are the ‘dynamic’ drivers of competitive versus cooperative
commercialisation?
• Under Schumpeterian competition, when is cooperative commercialisation
desirable?
Broad research agenda

• Builds on research program with Scott Stern (MIT) and David Hsu
(Wharton) ...

• Goal: use insights from bargaining theory to understand the


commercialisation choices of entrepreneurial start-ups and the impact of
these on the rate of innovation

• This paper ...


• What are the ‘dynamic’ drivers of competitive versus cooperative
commercialisation?
• Under Schumpeterian competition, when is cooperative commercialisation
desirable?

• Answering each requires a dynamic model and an environment where


competition for the market is more important than competition in the market.
Dynamic drivers of commercialisation choice
Dynamic drivers of commercialisation choice

• Static rationale for cooperative commercialisation is compelling


• Save duplicating complementary assets (Teece, 1986)
• Prevent rent dissipation (Gans & Stern, 2000)
Dynamic drivers of commercialisation choice

• Static rationale for cooperative commercialisation is compelling


• Save duplicating complementary assets (Teece, 1986)
• Prevent rent dissipation (Gans & Stern, 2000)
• But competition is often observed
• Arora and Gambardella (2009)
Dynamic drivers of commercialisation choice

• Static rationale for cooperative commercialisation is compelling


• Save duplicating complementary assets (Teece, 1986)
• Prevent rent dissipation (Gans & Stern, 2000)
• But competition is often observed
• Arora and Gambardella (2009)
• Competition can be partly explained by imperfections in ideas markets
• Theory: Arora, Anton & Yao
• Examples: Apple, Google, Genzyme, Intuit, Facebook ...
• Strong IP encourages cooperation: Gans, Hsu & Stern (2002); Arora & Ceccagnoli
(2005)
• Less IP uncertainty encourages cooperation: Gans, Hsu & Stern (2008)
Dynamic drivers of commercialisation choice

• Static rationale for cooperative commercialisation is compelling


• Save duplicating complementary assets (Teece, 1986)
• Prevent rent dissipation (Gans & Stern, 2000)
• But competition is often observed
• Arora and Gambardella (2009)
• Competition can be partly explained by imperfections in ideas markets
• Theory: Arora, Anton & Yao
• Examples: Apple, Google, Genzyme, Intuit, Facebook ...
• Strong IP encourages cooperation: Gans, Hsu & Stern (2002); Arora & Ceccagnoli
(2005)
• Less IP uncertainty encourages cooperation: Gans, Hsu & Stern (2008)
• But individual studies of start-up choice over commercialisation suggest a dynamic
concern for creation and capture of future innovative rents
• EMI (CT Scanner); Nucleon, Ecton and Palm
Antitrust and cooperative commercialisation
Antitrust and cooperative commercialisation

• Two schools
• ‘Hands off’: When there is competition for the market, there are no
competitive gains from intervention. Better not to disrupt innovation.
(Gilbert, Schmalensee, Evans, etc)
• ‘Vigilance’: In dynamic environments, it pays to be vigilant in promotion
competition so as not to lose future innovative potential
Antitrust and cooperative commercialisation

• Two schools
• ‘Hands off’: When there is competition for the market, there are no
competitive gains from intervention. Better not to disrupt innovation.
(Gilbert, Schmalensee, Evans, etc)
• ‘Vigilance’: In dynamic environments, it pays to be vigilant in promotion
competition so as not to lose future innovative potential

• Recent FTC/DOJ Draft Merger Guidelines


• “… effect is most likely to occur if at least one of the merging firms has
capabilities that are likely to lead it to develop new products in the future
that would capture substantial revenues from the other merging firm. The
Agencies therefore also consider whether a merger will diminish innovation
competition by combining two of a very small number of firms with the
strongest capabilities to successfully innovate in a specific direction.”
Starting point
Starting point

• Segal & Whinston (2007) model of dynamic technological competition


• SW apply to antitrust practices
• SW consider entrants as competing and displacing incumbents -- on-
going change
Starting point

• Segal & Whinston (2007) model of dynamic technological competition


• SW apply to antitrust practices
• SW consider entrants as competing and displacing incumbents -- on-
going change

• Here, modifications
• Remove static competition effects to focus on dynamics in
Schumpeterian, ‘winner-take-all’ or ‘for the market’ competition.
• Incorporate ideas markets
• Model dynamic capabilities (firms may not be long-lived)
Firms & Innovations
Firms & Innovations

• Discrete time, infinite horizon with common discount factor, δ


Firms & Innovations

• Discrete time, infinite horizon with common discount factor, δ

• Innovations are sequential


• Innovation yields new product of higher quality than previous generation
Firms & Innovations

• Discrete time, infinite horizon with common discount factor, δ

• Innovations are sequential


• Innovation yields new product of higher quality than previous generation

• Incumbent (I)
• There is a single producer of the new product who can earn per period
monopoly rents, Π, until displaced.
Firms & Innovations

• Discrete time, infinite horizon with common discount factor, δ

• Innovations are sequential


• Innovation yields new product of higher quality than previous generation

• Incumbent (I)
• There is a single producer of the new product who can earn per period
monopoly rents, Π, until displaced.

• Ideas are scarce


• For each product generation only one firm conducts R&D
• Innovation leader drawn from pool of firms including the current incumbent
(O’Donoghue, Scotchmer and Thisse, 1998)
• I can be the innovation leader and prolong incumbency
• Otherwise, label leader as ‘entrant’ or E.
Innovation Rate
Innovation Rate

• Paper: assume that an innovation leader generates an innovation immediately


Innovation Rate

• Paper: assume that an innovation leader generates an innovation immediately

• Here: for welfare focus on situations where we believe that incumbents


innovate less intensively than start-ups
• Cannabilisation (Arrow, 1962)
• Low powered incentives (Aghion and Tirole, 1994)
Innovation Rate

• Paper: assume that an innovation leader generates an innovation immediately

• Here: for welfare focus on situations where we believe that incumbents


innovate less intensively than start-ups
• Cannabilisation (Arrow, 1962)
• Low powered incentives (Aghion and Tirole, 1994)

• Key implication
• In welfare judgments, want policies that will encourage start-up rather than
incumbent innovation
Timeline

t-1 t t+1
Timeline

t-1 t t+1

Selection
Timeline

t-1 t t+1

Selection Production
Timeline

t-1 t t+1

Selection Production Innovation


Timeline

t-1 t t+1

Selection Production Innovation Negotiations


Timeline

t-1 t t+1

Selection Production Innovation Negotiations

Innovation
leader
chosen
Timeline

t-1 t t+1

Selection Production Innovation Negotiations

Innovation
leader Capabilities Acquired
chosen
Timeline

t-1 t t+1

Selection Production Innovation Negotiations

Innovation
Commercialisation
leader Capabilities Acquired
choice
chosen
Negotiations
Negotiations

Entrant
Innovates
Negotiations

Competition

Entrant
Innovates
Negotiations

I loses
Competition
monopoly
profits
Entrant
Innovates
Negotiations

I loses E
Competition
monopoly becomes I
profits (earns Π)
Entrant
Innovates
Negotiations

I loses E
Competition
monopoly becomes I
profits (earns Π)
Entrant
Innovates

Cooperation
Negotiations

I loses E
Competition
monopoly becomes I
profits (earns Π)
Entrant
Innovates

I retains
Cooperation monopoly
profits
Negotiations

I loses E
Competition
monopoly becomes I
profits (earns Π)
Entrant
Innovates

I retains E remains
Cooperation monopoly E
profits (earns τ)
Selection: Dynamic Capabilities
Selection: Dynamic Capabilities

Prior Activity
Selection: Dynamic Capabilities

Prior Activity

Production
Only
Selection: Dynamic Capabilities

Prior Activity

Production
Only

Innovation
Only
Selection: Dynamic Capabilities

Prior Activity

Production
Only

Innovation
Only

Production &
Innovation
Selection: Dynamic Capabilities

Prior Activity Example

Production
Only

Innovation
Only

Production &
Innovation
Selection: Dynamic Capabilities

Prior Activity Example

Production I not an
Only innovation leader

Innovation
Only

Production &
Innovation
Selection: Dynamic Capabilities

Prior Activity Example

Production I not an
Only innovation leader

Innovation E as innovation
Only leader

Production &
Innovation
Selection: Dynamic Capabilities

Prior Activity Example

Production I not an
Only innovation leader

Innovation E as innovation
Only leader

Production & Merger or I as


Innovation innovation leader
Selection: Dynamic Capabilities

Probability of
Prior Activity Example becoming Leader
Production I not an
Only innovation leader

Innovation E as innovation
Only leader

Production & Merger or I as


Innovation innovation leader
Selection: Dynamic Capabilities

Probability of
Prior Activity Example becoming Leader
Production I not an
σ p ∈[0,1]
Only innovation leader

Innovation E as innovation
Only leader

Production & Merger or I as


Innovation innovation leader
Selection: Dynamic Capabilities

Probability of
Prior Activity Example becoming Leader
Production I not an
σ p ∈[0,1]
Only innovation leader

Innovation E as innovation
leader
σ i ∈[0,1]
Only

Production & Merger or I as


Innovation innovation leader
Selection: Dynamic Capabilities

Probability of
Prior Activity Example becoming Leader
Production I not an
σ p ∈[0,1]
Only innovation leader

Innovation E as innovation
leader
σ i ∈[0,1]
Only

Production & Merger or I as


σ ip ≥ max{σ p , σ i }
Innovation innovation leader
Continuation payoffs
Continuation payoffs

I Innovation Leader
Continuation payoffs

I Innovation Leader
Continuation payoffs

I Innovation Leader

E Innovation Leader
Continuation payoffs

I Innovation Leader

E Innovation Leader
Continuation payoffs

I Innovation Leader

E Innovation Leader
Continuation payoffs

I Innovation Leader

E Innovation Leader

I Non-
Innovator
Continuation payoffs

I Innovation Leader

E Innovation Leader

I Non-
Innovator
Continuation payoffs

I Innovation Leader

E Innovation Leader

I Non-
Innovator
Continuation payoffs

I Innovation Leader

E Innovation Leader

I Non-
Innovator

Cooperation
Continuation payoffs

I Innovation Leader

E Innovation Leader

I Non-
Innovator

Cooperation Competition
Gains from Trade (Licensing)

Cooperation Competition
Gains from Trade (Licensing)

Cooperation Competition
Gains from Trade (Licensing)

+ +

Cooperation Competition
Gains from Trade (Licensing)

+ > +

Cooperation Competition
Gains from Trade (Licensing)

Π − τ + δσ pVI + δ (1 − σ p )v I + τ + δσ i v E
≥ Π + δσ p v E + δσ iVI + δ (1 − σ i )v I
Gains from Trade (Licensing)

( )
(σ p − σ i ) VI − v E − v I ≥ 0
Gains from Trade (Licensing)

(σ p − σ i )(σ ip − σ i − σ p ) ≥ 0
Gains from Trade (Licensing)

(σ p − σ i )(σ ip − σ i − σ p ) ≥ 0

Caveat: if incumbents innovate at a slower rate, both will have an interest


in choosing a commercialisation mode that maximises the probability that
the incumbent is the innovation leader
Licensing Equilibrium Outcomes
σ ip

Competition
VI < v I + v E

Licensing
σp Licensing

Competition
VI ≥ v I + v E

σi σ ip
Licensing Equilibrium Outcomes
σ ip

Competition
VI < v I + v E

Licensing
σp Licensing

Does high σi Competition


drive VI ≥ v I + v E
competition?
σi σ ip
Gains from Trade (Acquisition)
Gains from Trade (Acquisition)

Π − τ + σ ipδV + (1 − σ ip )δVI + τ
I
i

  
Joint Payoff from Cooperation

≥ σ pδVE + Π + σ iδV + (1 − σ i )δVI


I
i

  
Joint Payoff from Competition
Gains from Trade (Acquisition)

(σ ip − σ i )δ (VI − v I ) ≥ σ pδ v E
Gains from Trade (Acquisition)

(σ ip − σ i )δ (VI − v I ) ≥ σ pδ v E

1 − σ ip > 1 − σ i − σ p
Gains from Trade (Acquisition)

(σ ip − σ i )δ (VI − v I ) ≥ σ pδ v E

1 − σ ip > 1 − σ i − σ p implies positive externality


on potential entrants
Acquisition Equilibrium Outcomes
σ ip

VIi < vVII ++ vVEE


Competition

σp

Acquisition
Does high σi
drive VIi ≥ vVII ++ vVEE
competition?
σi σ ip
Welfare
Welfare

• Acquisition never socially desirable


Welfare

• Acquisition never socially desirable

• Significant: commentators suggest that Schumpeterian competition implies


loose merger policy
Welfare

• Acquisition never socially desirable

• Significant: commentators suggest that Schumpeterian competition implies


loose merger policy

• Here: merger consolidates capabilities to become next lead innovator in


incumbent’s hands (removes entrant capabilities) -- with reduced innovation
incentives.
Welfare

• Acquisition never socially desirable

• Significant: commentators suggest that Schumpeterian competition implies


loose merger policy

• Here: merger consolidates capabilities to become next lead innovator in


incumbent’s hands (removes entrant capabilities) -- with reduced innovation
incentives.

• Merger delays the forces of creative destruction


Alternative Contracting Possibilities
Alternative Contracting Possibilities

• Restructuring and spin-outs


• If incumbent innovation leader has lower chance of persisting, may prefer to spin-
out innovative element prior to innovation and production
• No gains from trade from licensing
Alternative Contracting Possibilities

• Restructuring and spin-outs


• If incumbent innovation leader has lower chance of persisting, may prefer to spin-
out innovative element prior to innovation and production
• No gains from trade from licensing
• Delayed negotiation
• If incumbent innovation leader has a higher chance of persisting, want to minimise
risk and delay licensing until after innovation leader is revealed or to renegotiate a
prior agreement upon the resolution of uncertainty
• However, this may, in reality, represent considerable cost in terms of delayed
commercialisation
Alternative Contracting Possibilities

• Restructuring and spin-outs


• If incumbent innovation leader has lower chance of persisting, may prefer to spin-
out innovative element prior to innovation and production
• No gains from trade from licensing
• Delayed negotiation
• If incumbent innovation leader has a higher chance of persisting, want to minimise
risk and delay licensing until after innovation leader is revealed or to renegotiate a
prior agreement upon the resolution of uncertainty
• However, this may, in reality, represent considerable cost in terms of delayed
commercialisation
• Entrant participation in production
• Wakeman (2008): one third of biotech start-up deals involve collaboration in
marketing, sales, clinical trials etc
• Increases domain for cooperation but causes firms to favour licensing over
acquisition
Conclusions
Conclusions

• Even in the absence of static rationales, dynamic considerations can drive


commercialisation choices
Conclusions

• Even in the absence of static rationales, dynamic considerations can drive


commercialisation choices

• Need to look at relative capabilities


• Capabilities here are based on experience
• Empirically, examine commercialisation choice based on start-up
experience in R&D variables and incumbent experience in products and
production
Conclusions

• Even in the absence of static rationales, dynamic considerations can drive


commercialisation choices

• Need to look at relative capabilities


• Capabilities here are based on experience
• Empirically, examine commercialisation choice based on start-up
experience in R&D variables and incumbent experience in products and
production

• Private commercialisation choice aimed at slowing down forces of creative


destruction
• Will cooperate if can reduce future innovation incentives
• But competition may also generate this outcome
• Difficult to find blanket policies against licensing (may want it to be
compulsory!) but acquisition is always welfare reducing

Você também pode gostar