Not every network is a monopoly: sulla concorrenza nei mercati online

dal saggio di uno dei maestri dell’antitrust (che si sta occupando molto dei mercati online)  Hovenkamp Herbert, Gatekeeper Competition Policy (February 4, 2023), intorno alla proposta di legge USA c.d. the American Innovation and Choice Online Act (AICOA)

<<While more studies need to be done, there is little reason today
for thinking that the exercise of market power is more common or
more harmful on online markets than on traditional markets.34 Internet
and traditional markets exhibit differing degrees of competition
depending on the product, including some monopoly. 35 Online firms
are more likely to be networked, and networking can be a source of
market power,36 but it can also result in better products, reduced costs,
or broader access.
Clearly networking can be a source of power because
networked markets appeal to a wider group of customers and can also
have lower costs. In addition to that are direct and indirect network
effects that can make networks appealing.
Not every network is a monopoly, however. Another
phenomenon that the internet facilitates is product differentiation, or
the assembly of different packages of products and other offerings.
For example, Facebook, Instagram, Twitter, LinkedIn, TikTok,
Reddit, and others are all social networking sites that are subject to
both direct and indirect network effects. Within a site, they become
more valuable as the number of users increases. They are not natural
monopolies or winner-take-all platforms, however, because of product
differentiation. The same thing can be said of countless newspapers
and other periodicals and dating sites, virtually all of which operate
mainly or exclusively online. (….)

With these realities, the best approach for antitrust policy is
some expansion of duties to deal that take network operational
obligations more seriously. But these rules should apply to every firm
that has substantial market power in a particular networked product or
service, not to a subset that is identified by absolute firm size, and then
without regard to power in a particular product. That approach is both
underinclusive as to firms and overinclusive as to products>>, pp. 10-12.