What are network effects?
A product displays positive network effects when more usage of the product by any user increases the product's value for otherusers (and sometimes all users). While formally different from and a more general concept than network externalities, much of the theory underlying network effects was developed to study network externalities, and the two terms are still used interchangeably. You can get a clear overview of the differences between the two concepts by reading Stan Liebowitz's Network Externalities (Effects). Other terms that refer to the same (or related) effects include positive externalities and demand-side economies of scale. Surveys of research on the economics of network effects can be found in Economides 1996 and Farrell and Klemperer 2001.
How common are network effects?
Network effects were first studied in the context of long-distance telephony in the early 1970's (one of the earliest papers on the topic is Rohlfs 1974). Today, they are widely recognized as a critical aspect of the industrial organization of IT industries, and are prevalent in a wide variety of sectors, including software, microprocessors, telecommunications, e-commerce and electronic marketplaces. Empirical evidence of network effects has been found in product categories as diverse as spreadsheets (Brynjolfsson and Kemerer 1996), databases (Gandal 1995), networking equipment (Forman 2001) and DVD players (Dranove and Gandal 2003)
A closer look at network effects
There are a more complex and subtle set of economics issues that may underly this simple characterization of network effects. A brief outline of some of them:
Direct network effects: The simplest network effects are direct: increases in usage lead to direct increases in value. The original example of telephone service is a good illustration of a product that displays direct network effects. This is the kind of network effect modeled by most work in this area, including the papers by Katz and Shapiro 1985 and Farrell and Saloner 1985, which many academics consider the most influential.
Indirect network effects: Network effects may also be indirect, where increased in usage of the product spawns the production of increasingly valuable complementary goods, and this results in an increase in the value of the original product. For instance, while there are some direct network effects associated with Windows (arising out of file compatibility), the indirect network effects that arise from the increased quality and availablity of complementary applications software are probably much more important. Economides and Salop 1992 provided one of the earliest frameworks and insight into the economics of indirect network effects; an interesting recent paper by Church, Gandal and Krause argues that indirect network effects also give rise to adoption externalities.
Two-sided network effects: Network effects can also be two-sided: increases in usage by one set of users increases the value of a complementary product to another distinct set of users, and vice versa. Both Rochet and Tirole 2001 and Armstrong 2002 provide accessible and clearly explained overviews. Hardware/software platforms, reader/writer software pairs, marketplaces and matching services display this kind of network effects. In many cases, one may think of indirect network effects as a one-directional version of two-sided network effects.
Local network effects: The microstructure of an underlying network of connections often influences how much network effects matter. For example, a good displays local network effects when rather than being influenced by an increase in the size of a product's user base in general, each consumer is influenced directly by the decisions of only a typically small subset of other consumers, for instance those he or she is "connected" to via an underlying social or business network (instant messaging is a great example of a product that displays local network effects). The extent of clustering in the network as well as the extent of information each customer possesses may become relevant in this context.
Here's an early paper that studied local network effects and complex network structure. There is now a growing literature studying more general network games. These new models draw actively from the 'science of networks', and a good place to start understanding this is by reading Mark Newman's remarkable survey on the structure and function of complex networks.
Compatibility and standards: In order for IT products to derive the benefits of network effects from each other, they need to be compatible. This often poses strategic trade-offs for firms, between the performance and backward-compatibility of evolving product lines, and between openness and control of core technologies. Moreover, ensuring the evolution of shared technology standards is critical in network industries, which can be difficult when competing technology firms each want their R&D to be well-represented.Shapiro and Varian's book have very insightful material on these issues; an interesting recent formal model of standard-setting is provided by Farrell 1996
A few economic consequences
When increases in usage cause an increase in value across all users, this creates a form of increasing returns, which changes the nature of competition substantially. Theories of competition in network industries emphasize the path dependence of outcomes, and suggest that early leads are important, intrinsically inferior products will frequently dominate superior products, and influencing customer expectations plays a crucial role in 'winning' in a network market. Brian Arthur's papers are particularly instructive on some of these issues.
These increasing returns often lead to an equilibrium in which a single firm or product dominates an industry segment. Witness Microsoft's near-total control over the office productivity and US desktop operating systems market, and the substantial profits they are able to generate from these product lines (combined operating income of over a billion dollars a month). Whether or not this is an efficient outcome was one of the secondary points of discussion in the recent Microsoft antitrust trial, and interesting perspectives were provided by, among others, Franklin Fisher, Richard Schmalensee and Nicholas Economides. Some of this discussion suggests that due to the persistent threat of entry, the equilibrium market structure of a network industry is often entry-deterring (rather than unfettered) monopoly, with the installed base of users serving as the deterrent; this idea has been modeled formally in Fudenberg and Tirole 2000; the effects of an entry threat on a network monopolist's ability to price-discriminate is studied in Sundararajan 2003.
Selected other links that I've found both interesting and useful
Brian Arthur's papers: Some of his widely read papers on increasing returns, path dependence and technological lock-in when firms compete in markets with network effects (and related characteristics).
The Economics of Networks: Nicholas Economides' site dedicated to network economics, the Microsoft case, telecommunications policy and related subjects. A remarkable collection of links to research papers on the subject.
Stan Liebowitz's page: Lots of information and links, as well as information about his books. If you're relatively new to the subject of network effects, I highly recommend his article Network externalities (effects)
The structure and function of complex networks: A great survey of the 'science of networks', covering topics that include random graphs, an excellent basis for modeling the structure and dynamics of networks, economic or otherwise. Written by Mark Newman, who is pretty prolific on the topic. Other interesting sites along the same lines include the Linked page and resources onscale-free and small-world networks from the International Network for Social Network Analysis.
The Tipping Point: Malcolm Gladwell's book about the dynamics of epidemics and related social phenomena. For anyone interested in delving deeper into the structural and dynamic aspects of network economics, reading this book is bound to generate ideas.
A product displays positive network effects when more usage of the product by any user increases the product's value for otherusers (and sometimes all users). While formally different from and a more general concept than network externalities, much of the theory underlying network effects was developed to study network externalities, and the two terms are still used interchangeably. You can get a clear overview of the differences between the two concepts by reading Stan Liebowitz's Network Externalities (Effects). Other terms that refer to the same (or related) effects include positive externalities and demand-side economies of scale. Surveys of research on the economics of network effects can be found in Economides 1996 and Farrell and Klemperer 2001.
How common are network effects?
Network effects were first studied in the context of long-distance telephony in the early 1970's (one of the earliest papers on the topic is Rohlfs 1974). Today, they are widely recognized as a critical aspect of the industrial organization of IT industries, and are prevalent in a wide variety of sectors, including software, microprocessors, telecommunications, e-commerce and electronic marketplaces. Empirical evidence of network effects has been found in product categories as diverse as spreadsheets (Brynjolfsson and Kemerer 1996), databases (Gandal 1995), networking equipment (Forman 2001) and DVD players (Dranove and Gandal 2003)
A closer look at network effects
There are a more complex and subtle set of economics issues that may underly this simple characterization of network effects. A brief outline of some of them:
Here's an early paper that studied local network effects and complex network structure. There is now a growing literature studying more general network games. These new models draw actively from the 'science of networks', and a good place to start understanding this is by reading Mark Newman's remarkable survey on the structure and function of complex networks.
A few economic consequences
When increases in usage cause an increase in value across all users, this creates a form of increasing returns, which changes the nature of competition substantially. Theories of competition in network industries emphasize the path dependence of outcomes, and suggest that early leads are important, intrinsically inferior products will frequently dominate superior products, and influencing customer expectations plays a crucial role in 'winning' in a network market. Brian Arthur's papers are particularly instructive on some of these issues.
These increasing returns often lead to an equilibrium in which a single firm or product dominates an industry segment. Witness Microsoft's near-total control over the office productivity and US desktop operating systems market, and the substantial profits they are able to generate from these product lines (combined operating income of over a billion dollars a month). Whether or not this is an efficient outcome was one of the secondary points of discussion in the recent Microsoft antitrust trial, and interesting perspectives were provided by, among others, Franklin Fisher, Richard Schmalensee and Nicholas Economides. Some of this discussion suggests that due to the persistent threat of entry, the equilibrium market structure of a network industry is often entry-deterring (rather than unfettered) monopoly, with the installed base of users serving as the deterrent; this idea has been modeled formally in Fudenberg and Tirole 2000; the effects of an entry threat on a network monopolist's ability to price-discriminate is studied in Sundararajan 2003.
Selected other links that I've found both interesting and useful
I was making my strategic management assignment on LinkedIn company. During research I have found this interesting article from NYU Stern (which I visited in March 2012). Quite interesting to see how network effect is affecting economics and business. SNA therefore is a great tool to exploit Network Effect possibilities.
Link to resource for educational purposes:
http://oz.stern.nyu.edu/io/network.html
1 comment:
Thoughtful, excellent post. Network effects are indeed one of the things that SNA can help measure more accurately than before
Post a Comment