Economie des rseaux, Economie de lInternet

1
Economie des réseaux, Economie de lInternet

• Nicolas Curien
• Ecole dété du GDR STIC
• 5 Septembre 2005

2
Deux Sujets Complémentaires

• Economie des réseaux étude du système
  concurrence-régulation dans les secteurs en
  réseau sujet déconomie industrielle
• Economie de lInternet étude de limpact des
  réseaux et services de communication électronique
  sur léconomie et la société sujet de
  prospective interdisciplinaire.

3
Quest-ce quun Réseau ?

• Réseau interconnexion (ingénieur) vs Réseau
  intermédiation (économiste)
• Morphologie en couches infrastructure,
   infostructure  et services
• Critères de reconnaissance
• Effets de club
• Synergies de coûts
• Subventions croisées
• Partage monopole / marché
• Dispositifs de régulation

4
Réseaux, Clubs de Consommation

• Biens réseaux et biens-systèmes
• Modélisation des effets de club effet de masse
  critique
• Dynamique de croissance dun réseau (Ex. Le
  téléphone en France)
• Guerres de réseaux émergence dun standard
• Modèles de diffusion technologique (déterministes
  et évolutionnistes)

5
Réseaux, Systèmes de Production

• Synergies de coûts effets déchelle,
  denvergure, de  réseau 
• Théories du monopole naturel et de la
  contestabilité
• Régulation des  facilités essentielles  en
  présence dasymétries dinformation
• Tarifications de second rang en présence
  dexternalités de club et de contraintes de
  capacité
• Subventions croisées et soutenabilité

6
Réseaux, Concurrence et Régulation

• Couches morphologiques et concurrence  Cône
  douverture  du marché
• 3 Axes de la  déréglementation 
• Libéralisation du marché
• Evolution de la régulation
• Evolution de lopérateur historique
• Matrice  Marché x Régulation 
• Régulation sectorielle vs Droit de la
  concurrence, dynamique de la régulation

7
Chantiers de la Régulation

• Licences et autorisations soumission
  comparative vs enchères (UMTS, BLR)
• Interconnexion (téléphone) ECPR / CMILT
• Accès (haut débit) Modèle de  léchelle des
  investissements 
• Service public et service universel Modèles
  dévaluation de coût
• Processus dAnalyse des marchés la matrice
   Services x Marchés 

8
Dun Sujet to Another Topic

• Au delà de léconomie interne des réseaux et
  services de communication électroniques .
• What are the expected impact of their diffusion
  on economic and social mechanisms ?
• From Internet as a  Network  to Internet as a
   Laboratory 

9
Sketchy Overview

• The utopia of the  New Economy  (1995-2000)
  misinterpreted the actual consequences of the
   Digital Revolution .
• The economic value will originate in
   knowledge  rather than in  information .
  Knowledge economy society is much better a
  phrasing than Information economy society.
• After Solows productivity paradox (1980), a
  second paradox now arises the knowledge economy
  will be a hybrid one, gathering both features of
  a  market economy  and of a  public economy .
• Online Communities constitute the key  informal
  institution  of the knowledge economy.

10
2 Key-Drivers of the Digital Revolution 

• Digitalization of information
• Cultural goods (books, music, movies)
• Software
• Video-games
• Information intensive goods
• Experience goods
• Attention goods
• Complex goods
• Innovating goods

11
2 Major Economic Impacts

• The emergence of the informational commons 
• Information becomes a  non-rival  good,
• despite of the current industrial and
  governmental attempts to protect non rival but at
  least partially  excludable  contents
  (eradication of P2P seemingly to protect property
  rights. and mostly to protect unduly the
  business of majors).
• The request for  meta-information 
• In a complexified space of goods and services,
  economic agents need higher compentecies
• more off-market  and/or  off-hierarchy 
  information is necessary in order to supply, to
  innovate or to consume.

12
The mistakings of the New Economy 

• The initial simplistic and utopic view
• ICTs will tend to make
• markets more fluid (B2B B2C aiming at a perfect
  market),
• hierarchies more controlable (aiming at a perfect
  bureaucracy),
• thus
• generating increased efficiency and productivity
  gains,
• without major changes in the organisation of
  economy, except for welfare improvement.
• What really occurs
• more segmented markets,
•  flatter  hierarchies with more elastic and
  permeable frontiers,
• in depth transformation of many business models
  (especially contents industries and information
  intensive industries),
• Schumpeterian  destructive creation  making the
  assessment of welfare impacts uneasy.

13
Economy of Knowledge 

•  Codified  information is a free input rather
  than a valuable output
• digitalization indeed creates value as it
  generates productivity gains,
• but as digitalized information becomes a non
  rival good supplied at almost zero marginal cost,
• then optimal price approaches zero, so that
  economic value no longer can be collected at this
  stage.
• A new type of scarcity originates in the
   tacit  information necessary for
• matching supply and demand (info-mediation,
  Hayekian market),
• coupling innovation to usersneeds (as in
  open-source),
• transforming information into knowledge at both
  the individual level (learning process) and the
  supra-individual level (generation of management
  routines).

14
3 Main Characteristics

• Open model
• Producers and consumers participate in the same
   social algorithm ,
• where info-mediation replaces the traditional
  direct hierarchy / market interface,
• the consumer becoming active as a tester or even
  as a co-producer,
• which leads to a better fit of  rational 
  production and  hedonic  consumption
• in a world of fast innovation.
• Circulating goods
• The repeated usage of informational goods does
  not destroy but creates value,
• so that the free circulation of those goods
  benefits the collectivity,
• that circulation being globally by itself a non
  rival asset,
• even if some of the circulating goods
  individually are (or could be made) rival (just
  as the Kula objects exchanged by Papouasian
  tribes).
• Consumption spillovers
• Consumers benefit from critics and advices
  supplied by other consumers,
• ex ante when making buy decicions,
• ex post when learning how to use goods.

15
After Solow a 2nd Paradox

• Considered as a technological  neutral 
  platform
• ICTs seem to reinforce the  market economy  and
  world-wide competition,
• by enhancing transparency, universality,
  flexibility, fluidity of transactions.
• Considered as the vector of the digitalization of
  contents
• ICTs generate the typical characteristics of a
   public economy , such as
• strong economies of scale (fixed cost economy),
• powerful club effects (created by electronic
  networks and standards),
• informational  commons .
• favouring concentration (Microsoft) and/or
  cooperation (RD, B2B monopsonies), rather than
  unrestricted competition (model of
   coopetition ).
• The actual  new economy  will be a  hybrid
  economy ,
• mixing the features of the two opposite models
  (dialectic rather than dichotomic evolution!),
• relying upon an original kind of informal
  institutional framework online communities
  (OCs).

16
Online Communities (OCs)

• OCs are  endogenous , spontaneous and informal
  institutions generating a new model of
  inter-individual interaction which extends the
  traditional models of social networks (SNs) and
  mass medias (MMs), and partially substitute to
  those.
• OCs are of various types
• Practice OCs (final markets, ex post)
• Experience OCs (final markets, ex ante)
• Epistemic OCs (innovation)
• Inter-firm and professional OCs (intermediary
  markets)
• Intra-firm OCs (hierarchies)
• However OCs share common features that oppose
  them to both MMs and SNs.

17
3 Infomediation Structures

• Social Networks (SNs)
• Information circulates along a spatialized graph
  structure showing vicinity links, distant links
  and clusters.
• Strength of weak links ( small world  according
  to Granovetter)
• Attempts to densify SNs through selective (or
  even  secret ). communities (civil servants
  corporations in France, macons).
• Mass Medias (MMs)
• One way information flow (point to many) or
  two-step flow (MM SN)
• Poor differenciation of messages (narrow
  selection of products).
• Almost no feedback (except for statistical
  information about audience).
• Online Communities (OCs)
• Almost no direct inter-individual relationships
  accross participants.
• Indirect relationships through the informational
  corpus ( blackboard ).
• Reverse free-riding  writing  on the
  blackboard exceeds  reading , requesting
  quality rating.

18
Infomediation Graphs
19
3 Types of Social Link

• Social Networks (SNs)
• Symmetric roles of agents
• Long run intimacy
• Links do survive temporary inactivity
• Tacit reciprocal obligations
• Separate networks depending on the context (work
  / home)
• Mass Medias (MMs)
• Strongly asymmetric roles of the source and the
  audience
• Short run and/or fictitious intimacy (star / fan)
• Transfer of attention (TV programs attract
  audience for advertisement)
• Online Communities (OCs)
• Complementary dedicated roles authors or
  innovators /experts / novices
• Instrumental intimacy (ephemerous personalized)
• Asynchronous relationships mediated through the
  blackboard
• Durable but fragile link to the blackboard

20
Instrumental Intimacy
21
From Code to Semantics

• In both cases of social networks (SNs) and mass
  medias (MMs), information may be considered as a
  signal
• Meaning of the message is rather easy to decode,
  whereas trust in its truth and authenticity is at
  stake.
• Signalling is a  natural  (innate) phenomenon
  (ecology)
• The economics of information is  Shannonian  as
  it treats information as a coded signal
  (principal/agent model, perfect bayesian
  equilibria in game theory)
• In the case of Online Communities (OCs),
  information acquires a semantic dimension as in a
  language
• Trust is not a major issue (it may be solved
  endogeneously), whereas the very meaning of
  messages is costly to decode (because of fuzzy
  formulation).
• Language is an  artificial  (acquired)
  construct.
• The coding is ambiguous and of a symbolic type
  (links between objects rather than objects
  themselves are encoded, Savage Thinking, Levi
  Strauss).
• Meta-representation (especially representation of
  others ignorance) is essential to the
  interactive process of collaborative construction
  of sense.

22
Comparative Performances of OCs, SNs and MMs

• Hierarchies and intermediary markets (OCs vs.
  SNs)
• Better decentralization of decisions (through
  information circulation at the lowest level) ?
• Better coupling accross hierarchies (Do OCs
  perform better than inter-personnal networks) ?
• Final markets (OCs vs MMs SNs)
• Impact on competition and product differenciation
  (same degree of competition with higher
  diffenrenciation) ?
• Better consumption of experience goods ?
• Better learning of how to use complex goods ?
• Better innovation through interactive interplay
  of innovators and users ?

23
Brief Summary

• Final markets are more and more  assisted 
  (self-regulated) by experience OCs, that tend to
  substitute for MMs and to extend SNs (word of
  mouth).
• Hierarchies and intermediary markets are more and
  more challenged by work-practice OCs and
  transversal professional OCs, that tend to move
  the firms boundary and alter the control of
  proprietary knowledge.
• The consumers production function (Lancaster
  model) develops, thus transforming the final
  market into an intermediary market and requesting
  the emergence of domestic-practice OCs.
• Despite of their diversity (experience, practice,
  epistemic, files sharing, etc.), OCs are the
  meeting point of  rationalized  production and
   savage  consumption (in Levi Strauss
  understanding of the term).
• The different types of OCs do share a same
  pattern of original features, in terms of
  structure and mechanisms, that oppose them to
  both SNs and MMs and raises the delicate issue of
  performance assessment.
• OCs are a new informal institution at the core of
  the K-economy society.