Economics of Network Industries
Also study required readings (*)
Topic 1: Introduction to network industries
1.1 Examples and scope of network industries
Network industries are industries where the value of a product or service changes with the number
of users of the same or compatible product or service.
• Compatibility matters: “same or compatible” hints at how standards can unify or fragment
markets.
• Network effects modify demand: a user’s willingness to pay increases with total adoption. This
breaks standard downward-sloping demand assumptions.
Network externality is the impact that an additional user of a product or service, or an additional
participant to some interaction, has on the value that other users or participants attach to this
product, service or interaction.
® Additional users increase value for existing users
® Ex. Messeging apps, social media apps, marketplaces
We will see that this leads to different market outcomes and policy issues compared to industries
without network externalities.
Network industries are closely related to platforms:
• Entities that bring together economic agents and actively manage network effects between
them
• Platforms have become increasingly dominant in the economy (and are still becoming more
important)
NETWORK INDUSTRIES IN THE WORLD ECONOMY
Many physical industries are network industries
- Telecommunication & internet infrastructure
- Radio, television, and news
- Financial networks (ex. Stock market)
- Electricity
- Transportation networks: Railroads, roads, airlines
Most of digital life is organizated on platforms. Some industries experience network externalities
through complementary goods and services
- Hardware/software ecosystems
- Printers and ink
- Game consoles and compatible video games
- Mobile OS/app stores
This leads us to two types of networks:
1. Direct network effects (e.g., phones, social networks)
2. Indirect (cross-side) network effects from complements (e.g., consoles ↔ games, OS ↔
apps)
Many of the high-growth digital enterprises of the 21th century are platforms or in network industries.
- Social media platforms: the value depends strongly on the number of users (within-group
network effects) ® Tiktok, Youtube, Facebook,…
, - Online marketplaces: connect buyers and sellers (cross-group network effects) ® Amazon,
Walmart, eBay,…
Unlike social media (within-group effects), marketplaces depend on complementary groups.
Realize how this changes competition dynamics: attackers need to recruit both sides
simultaneously (the classic chicken-and-egg problem).
- Search engine: the more users, the better the service of becomes and the more attractive for
advertisers. (within-group and cross-group network effects) ® Google
- Cloud computing: the more users, the better infrastructure and integration by developers and
businesses. ® Amazon, Microsoft, Google, Alibaba
Network industies are not new. Economic platforms existed long before digital technology. The core
function is reducing transaction costs and enabling interactions. Technology only changes the scale
and speed. Medieval trade fairs functioned as early platforms: central locations where traders could
meet under guaranteed rules and security.
STRATEGIES AND PUBLIC POLICY IN NETWORK INDUSTRIES
Some market structures are more concentrated than others. For example, the market structure for
smartphone operating systems is much more concentrated than the market structure for
smartphones. This market structure concentrated over time. ® Role of network effects
® Why are network industries often highly concentrated?
® How are competitive strategies different?
® When should regulators intervene?
Digital markets challenge traditional competition law because harm may occur without immediate
price effects (e.g., reduced innovation or choice). Tools like the EU’s Digital Markets Act (DMA) and
“gatekeeper” regulation respond to these concerns.
Possible interventions:
- Price controls or subsidies
- Technical compatibility mandates
- Competition policy: break-ups, preventing killer acquisitions, regulating gatekeepers.
BUSINESS MODELS IN THE DIGITAL ECONOMY
Digital business models
frequently use zero-price
strategies (Facebook,
Google Search).
“Free” services are cross-
subsidized by advertisers or
data monetization.
This slide connects business
models to multi-sided
markets, which will be
studied later.
1.2 Key economic characteristics of network industries
There are 4 main characteristics of network industries, these are common but are not always present
to the same degree or in the same form.
,Four core features:
1. Network externalities
2. Complementarity, compatibility and standardization
3. Significant economies of scale
4. Switching costs and lock-in
These four features collectively explain concentration, market power, and policy interest. They also
explain why digital markets evolve differently than classic markets: high fixed cost, low variable cost,
and feedback loops create strong path dependence.
1) NETWORK EXTERNALITIES
The consumption of a good involves network externalities if the utility depends on the number of other
people using the same or compatible product.
® Nobody would go on a social media platform with nobody else on it, nobody would buy a game
console without any games on it.
Metcalfe’s Law:
• The total value of a network of size n is
proportional to n(n – 1) due to
communication possibilities of n users.
• So the marginal value of a network is
proportional to number of users n.
The market outcome depends on consumers’ expectations on the number of users
- Expectations are crucial: people buy what they believe others will buy.
- Multiple equilibria are therefore possible (adoption vs. non-adoption, multiple standards).
- This explains why unsuccessful products sometimes are technologically superior but lose
because of coordination problems.
- Network externalities create S-shaped adoption curves: slow → fast → saturation.
For instance, the fax was invented in the 1950s, but only adopted between 1982 and 1987.
2) COMPLEMENTARITY, COMPATIBILITY AND STANDARDS
There is complementarity between two products if utility of consuming them together is larger than
utility of consuming them separately (opposite of substitutability)
® To work together as complements, products must be compatible
® To be compatible, products must work on the same standard.
® This creates coordination problem: risk to standardize on inferior technology
Ex.Computers are useful only with monitors & software, Game consoles only useful with games, etc.
Format wars = competition between standards
® Competition between standards (or systems) may or may not lead to “tipping”: one standard
becomes dominant.
® Open systems grow ecosystems faster, while closed systems may have better quality but
struggle to reach critical mass.
® Network effects can outweigh technological superiority.
Ex. Betamax vs VHS. Betamax was released by Sony in the mid 70s, VHS by JVC shortly after. Betamax was
technologically superior (picture, audio, size, …) but Sony applied a restrictive IP approach, and JVC freely
licensed VHS to manufacturers
This created network effects and drove down VHS prices, leading to VHS becoming the eventual standard. Sony
started producing VHS in 1988
Ex. IOS vs Android did not lead to tipping
Ex. Playstation, Nintendo, Xbox games did not lead to tipping
, 3) SIGNIFICANT ECONOMIES OF SCALE
Production of the first copy of software (or any information product) involves a huge sunk cost (which
cannot be recovered), whereas any further copy costs almost nothing to reproduce.
Ex. The cost to redevelop the Linux kernel version 2.6.0 in a traditional proprietary development setting has been
estimated to be $612 million USD, while it costs $0 to make a copy.
A cost function 𝐶(𝑞) exhibits economies of
scale when the average cost decreases as
output increases:
𝐶(𝑞)
𝐴𝐶(𝑞) = 𝑑𝑒𝑐𝑙𝑖𝑛𝑒𝑠 𝑤𝑖𝑡ℎ 𝑞
𝑞
This is typical for digital and information goods.
Most costs are upfront fixed costs
(development, infrastructure), while the
marginal cost of producing an additional unit is
very small or close to zero.
A common example:
𝐹
𝐶(𝑞) = 𝐹 + 𝑐𝑞 ⇒ 𝐴𝐶 (𝑞) = + 𝑐
𝑞
Here:
- 𝐹: large fixed (sunk) cost
- 𝑐: very low marginal cost
!
- As 𝑞 increases, " becomes small → average cost declines
Implications for market structure:
1. A competitive equilibrium cannot exist
o Under perfect competition, price = marginal cost = 𝑐.
o But 𝑝 = 𝑐 < 𝐴𝐶(𝑞), so firms cannot recover the fixed cost 𝐹.
o Therefore, competitive pricing drives firms out of the market.
2. Only a monopolist (or very few large firms) can survive
o A monopolist can set a monopoly price 𝑝# above marginal cost to cover fixed costs.
o This gives rise to natural monopoly properties: one firm can serve the entire market
more efficiently than many firms.
Connection to digital markets:
1. Global scale with very low operating cost: Once the fixed cost is paid (software development,
cloud infrastructure), distributing additional units - downloads, queries, API calls - is almost
free.
2. High concentration: Economies of scale + network effects reinforce each other:
o low marginal cost → large firms are advantaged
o network effects → users gravitate to the biggest network
Together, they lead to winner-takes-most markets.
3. Zero-price products become rational: Since marginal cost ≈ 0, giving away the product (e.g.,
Google Search, Facebook) is economically sensible.
Firms then monetize via: advertising, data, complementary services (e.g., cloud, premium,…)
REMINDER: Perfect competition P = MC
- Assumptions (network industries violate multiple):
o Many buyers and sellers
o Identical (homogeneous) products
o Buyers and sellers have perfect information
o No barriers to either entry or exit
Also study required readings (*)
Topic 1: Introduction to network industries
1.1 Examples and scope of network industries
Network industries are industries where the value of a product or service changes with the number
of users of the same or compatible product or service.
• Compatibility matters: “same or compatible” hints at how standards can unify or fragment
markets.
• Network effects modify demand: a user’s willingness to pay increases with total adoption. This
breaks standard downward-sloping demand assumptions.
Network externality is the impact that an additional user of a product or service, or an additional
participant to some interaction, has on the value that other users or participants attach to this
product, service or interaction.
® Additional users increase value for existing users
® Ex. Messeging apps, social media apps, marketplaces
We will see that this leads to different market outcomes and policy issues compared to industries
without network externalities.
Network industries are closely related to platforms:
• Entities that bring together economic agents and actively manage network effects between
them
• Platforms have become increasingly dominant in the economy (and are still becoming more
important)
NETWORK INDUSTRIES IN THE WORLD ECONOMY
Many physical industries are network industries
- Telecommunication & internet infrastructure
- Radio, television, and news
- Financial networks (ex. Stock market)
- Electricity
- Transportation networks: Railroads, roads, airlines
Most of digital life is organizated on platforms. Some industries experience network externalities
through complementary goods and services
- Hardware/software ecosystems
- Printers and ink
- Game consoles and compatible video games
- Mobile OS/app stores
This leads us to two types of networks:
1. Direct network effects (e.g., phones, social networks)
2. Indirect (cross-side) network effects from complements (e.g., consoles ↔ games, OS ↔
apps)
Many of the high-growth digital enterprises of the 21th century are platforms or in network industries.
- Social media platforms: the value depends strongly on the number of users (within-group
network effects) ® Tiktok, Youtube, Facebook,…
, - Online marketplaces: connect buyers and sellers (cross-group network effects) ® Amazon,
Walmart, eBay,…
Unlike social media (within-group effects), marketplaces depend on complementary groups.
Realize how this changes competition dynamics: attackers need to recruit both sides
simultaneously (the classic chicken-and-egg problem).
- Search engine: the more users, the better the service of becomes and the more attractive for
advertisers. (within-group and cross-group network effects) ® Google
- Cloud computing: the more users, the better infrastructure and integration by developers and
businesses. ® Amazon, Microsoft, Google, Alibaba
Network industies are not new. Economic platforms existed long before digital technology. The core
function is reducing transaction costs and enabling interactions. Technology only changes the scale
and speed. Medieval trade fairs functioned as early platforms: central locations where traders could
meet under guaranteed rules and security.
STRATEGIES AND PUBLIC POLICY IN NETWORK INDUSTRIES
Some market structures are more concentrated than others. For example, the market structure for
smartphone operating systems is much more concentrated than the market structure for
smartphones. This market structure concentrated over time. ® Role of network effects
® Why are network industries often highly concentrated?
® How are competitive strategies different?
® When should regulators intervene?
Digital markets challenge traditional competition law because harm may occur without immediate
price effects (e.g., reduced innovation or choice). Tools like the EU’s Digital Markets Act (DMA) and
“gatekeeper” regulation respond to these concerns.
Possible interventions:
- Price controls or subsidies
- Technical compatibility mandates
- Competition policy: break-ups, preventing killer acquisitions, regulating gatekeepers.
BUSINESS MODELS IN THE DIGITAL ECONOMY
Digital business models
frequently use zero-price
strategies (Facebook,
Google Search).
“Free” services are cross-
subsidized by advertisers or
data monetization.
This slide connects business
models to multi-sided
markets, which will be
studied later.
1.2 Key economic characteristics of network industries
There are 4 main characteristics of network industries, these are common but are not always present
to the same degree or in the same form.
,Four core features:
1. Network externalities
2. Complementarity, compatibility and standardization
3. Significant economies of scale
4. Switching costs and lock-in
These four features collectively explain concentration, market power, and policy interest. They also
explain why digital markets evolve differently than classic markets: high fixed cost, low variable cost,
and feedback loops create strong path dependence.
1) NETWORK EXTERNALITIES
The consumption of a good involves network externalities if the utility depends on the number of other
people using the same or compatible product.
® Nobody would go on a social media platform with nobody else on it, nobody would buy a game
console without any games on it.
Metcalfe’s Law:
• The total value of a network of size n is
proportional to n(n – 1) due to
communication possibilities of n users.
• So the marginal value of a network is
proportional to number of users n.
The market outcome depends on consumers’ expectations on the number of users
- Expectations are crucial: people buy what they believe others will buy.
- Multiple equilibria are therefore possible (adoption vs. non-adoption, multiple standards).
- This explains why unsuccessful products sometimes are technologically superior but lose
because of coordination problems.
- Network externalities create S-shaped adoption curves: slow → fast → saturation.
For instance, the fax was invented in the 1950s, but only adopted between 1982 and 1987.
2) COMPLEMENTARITY, COMPATIBILITY AND STANDARDS
There is complementarity between two products if utility of consuming them together is larger than
utility of consuming them separately (opposite of substitutability)
® To work together as complements, products must be compatible
® To be compatible, products must work on the same standard.
® This creates coordination problem: risk to standardize on inferior technology
Ex.Computers are useful only with monitors & software, Game consoles only useful with games, etc.
Format wars = competition between standards
® Competition between standards (or systems) may or may not lead to “tipping”: one standard
becomes dominant.
® Open systems grow ecosystems faster, while closed systems may have better quality but
struggle to reach critical mass.
® Network effects can outweigh technological superiority.
Ex. Betamax vs VHS. Betamax was released by Sony in the mid 70s, VHS by JVC shortly after. Betamax was
technologically superior (picture, audio, size, …) but Sony applied a restrictive IP approach, and JVC freely
licensed VHS to manufacturers
This created network effects and drove down VHS prices, leading to VHS becoming the eventual standard. Sony
started producing VHS in 1988
Ex. IOS vs Android did not lead to tipping
Ex. Playstation, Nintendo, Xbox games did not lead to tipping
, 3) SIGNIFICANT ECONOMIES OF SCALE
Production of the first copy of software (or any information product) involves a huge sunk cost (which
cannot be recovered), whereas any further copy costs almost nothing to reproduce.
Ex. The cost to redevelop the Linux kernel version 2.6.0 in a traditional proprietary development setting has been
estimated to be $612 million USD, while it costs $0 to make a copy.
A cost function 𝐶(𝑞) exhibits economies of
scale when the average cost decreases as
output increases:
𝐶(𝑞)
𝐴𝐶(𝑞) = 𝑑𝑒𝑐𝑙𝑖𝑛𝑒𝑠 𝑤𝑖𝑡ℎ 𝑞
𝑞
This is typical for digital and information goods.
Most costs are upfront fixed costs
(development, infrastructure), while the
marginal cost of producing an additional unit is
very small or close to zero.
A common example:
𝐹
𝐶(𝑞) = 𝐹 + 𝑐𝑞 ⇒ 𝐴𝐶 (𝑞) = + 𝑐
𝑞
Here:
- 𝐹: large fixed (sunk) cost
- 𝑐: very low marginal cost
!
- As 𝑞 increases, " becomes small → average cost declines
Implications for market structure:
1. A competitive equilibrium cannot exist
o Under perfect competition, price = marginal cost = 𝑐.
o But 𝑝 = 𝑐 < 𝐴𝐶(𝑞), so firms cannot recover the fixed cost 𝐹.
o Therefore, competitive pricing drives firms out of the market.
2. Only a monopolist (or very few large firms) can survive
o A monopolist can set a monopoly price 𝑝# above marginal cost to cover fixed costs.
o This gives rise to natural monopoly properties: one firm can serve the entire market
more efficiently than many firms.
Connection to digital markets:
1. Global scale with very low operating cost: Once the fixed cost is paid (software development,
cloud infrastructure), distributing additional units - downloads, queries, API calls - is almost
free.
2. High concentration: Economies of scale + network effects reinforce each other:
o low marginal cost → large firms are advantaged
o network effects → users gravitate to the biggest network
Together, they lead to winner-takes-most markets.
3. Zero-price products become rational: Since marginal cost ≈ 0, giving away the product (e.g.,
Google Search, Facebook) is economically sensible.
Firms then monetize via: advertising, data, complementary services (e.g., cloud, premium,…)
REMINDER: Perfect competition P = MC
- Assumptions (network industries violate multiple):
o Many buyers and sellers
o Identical (homogeneous) products
o Buyers and sellers have perfect information
o No barriers to either entry or exit
