page image

CHAPTER III - Network Power

Network Power
If networks are a new source of power, then it is important to understand how they work. What are the principles that govern network behavior? What are the key characteristics of networks that determine how and when leverage can (and cannot) be applied?

Tom Wheeler, Visiting Fellow at the Brookings Institution and former Chairman of the Federal Communications Commission (FCC), started the discussion proposing that “we look like our networks” and that how we connect with each other defines our economic and social activities. To illustrate the point, he looked back to the 19th century to the advent, first, of high speed transportation networks in the form of transcontinental railroads, and then the emergence of high-speed electronic communication networks, beginning with the telegraph. These two innovations “created the urban maps that we know today” as well as new economic and social structures such as mass production, the factory model of education and our modern health care system. Big government arose to deal with these forces, helping to reinforce the concept of the nation-state.

Now, 150 years later, we are witnessing a shift from centralizing networks to decentralizing networks, from centripetal to centrifugal forces that seem to be pushing power and activity outwards. ut networks have a Janus-like quality: even as they erode the power of traditional institutions by decentralizing control, they are also creating new forms of control. While we are familiar with the concept of “gatekeepers” who controlled access to mass media (publishers, editors, producers), we are now seeing the rise of “infokeepers” who understand the power of collecting and analyzing data, and “mathkeepers” who create the algorithms that determine much of what happens online. We are also redefining what we mean by a “network.” In addition to being just a collection of nodes and links, a sufficiently pervasive network can also be a platform on which other sub-networks can be built using the platform’s standards (think of the Internet itself, or a dominant application such as Google or Facebook).

These new roles and new configurations of power represent challenges to existing rules and assumptions that were developed in a different environment. For example, what is the relevance of anti-trust laws in an era where the issue is not just the production of goods and services but also the creation and manipulation of data? More broadly, how do we understand a society that is shifting from mass audiences (created and shaped by mass media) to a constellation of smaller and rapidly shifting networked communities?

Wheeler concluded by noting that what is most transformative and consequential about new innovations is not the underlying technology itself, but rather the secondary effects from that technology. Again, this is not new. In the 19th century, the mass urbanization created by the new transportation and communication networks gave rise to urban squalor that led to the development of zoning laws and public health services, while the growth of factories and mass production led to the creation of unions for workers. The question today is: what responses do we need to pursue to deal with the changes brought about by the new network technologies and their secondary effects?

Network roles. Networks come in many different forms and serve many different types of purposes. But there are some fundamental principles about how networks operate that are useful to keep in mind.

As noted earlier, the basic definition of a network is a collection of nodes and links. While power in hierarchies flows from the top, power in (horizontal) networks derives from the centrality of a node, which expresses how well connected and therefore important a given node is in a network. Since network participation is generally about influence rather than control, centrality, which can be measured in multiple ways, is a key factor.

Another important characteristic of networks is their degree of openness. At one extreme are tribes (like those that the young people that Marc Nathanson and Madeleine Albright met in Kenya belonged to), where membership generally requires that one be born into it. Tribal connectivity is typically maintained through multiple dimensions, including language, religion, dress and other customs. At the other extreme is the Internet or Facebook, which anyone can join and use (as some two billion people have) as long as they conform to a certain set of standards or protocols.

Scale is another critical dimension of networks. Networks tend to grow according to a power law, which means that their value increases exponentially with the number of participants, which in turn fuels further growth. With more than two billion users worldwide, Facebook functions as a “platform” (Esther Dyson) or a “network of networks” (Monika Bickert) that makes it easy for groups to form their own sub-communities.

Perhaps the most important question is what, exactly, are networks good for? In The Chessboard & the Web, Anne-Marie Slaughter identifies three basic types of networks, each with three subtypes, based on their primary function—networks that build resilience, networks that help groups to carry out discrete tasks, and networks that enable large groups to enlarge the impact of effective solutions (see sidebar, “A Network Typology”). In every case, key considerations include not only how the network functions technically but what the purpose of the network is, who it connects, and how participants communicate and work with each other through the network. Some networks are relatively ad hoc and transient, assembled to accomplish a specific goal, such as coordinating responses to an immediate crisis, then dissolving when the goal is reached. Other networks are more persistent because they serve an enduring need or purpose.

A Network Typology

  1. Resilience Networks
    The function of these strategic networks is to help to “avoid or respond to a crisis” in the short-, medium- or longer-term.
    1. Defense Networks enable an effective immediate response to a direct attack. Because they need to operate in extreme circumstances, basic survivability is a key requirement. The classic example is a packet-switched network, first described by Paul Baran of RAND in 1968 as a means to create a radically decentralized network that would be less vulnerable to a military attack than the centralized circuit-switched networks of the day. Baran’s innovation provided the technical basis for the Internet, which is well on the way to entirely replacing older network designs globally. A contemporary example of a defense network is the effort by the Department of Homeland Security to create “a distributed grid to improve cybersecurity” through real-time sharing information on cyberthreats and countermeasures.
    2. Response Networks are designed to coordinate ongoing responses to disasters and other emergencies. Ushahidi, which was originally created in 2007 as a network tool to document election fraud in Kenya has been used to coordinate responses to disasters such as the 2010 earthquake in Haiti.
    3. Stabilization Networks are intended to aid recovery from a disaster (including war) over time. Even as traditional activities such as capacity building are carried on, these networks help to strengthen connections between people who can rebuild fragmented communities.
  1. Task Networks
    These networks are intended to help disparate groups of participants to pursue specific projects.
    1. Cooperation networks support groups working to carry out a prescribed task in a prescribed manner. In an international context, such networks can build trust, even among adversaries, by connecting people to work together to accomplish a shared goal.
    2. Collaboration networks bring people together to figure out how to accomplish a common goal. A powerful example is the use by General Stanley McChrystal of a daily teleconference to share critical information among special operations units fighting in Iraq and Afghanistan as a tool to meld these highly independent units into a more effective “team of teams” by building a sense of a shared mission.
    3. Innovation networks expand the scope of people who are working to generate new ideas or find new approaches to problems. InnoCentive is a “crowdsourcing platform” that allows groups to post open “challenges” and offer a reward for the best solution. Interestingly, many InnoCentive winners have lacked conventional credentials that would have made them logical candidates to work on the problems they addressed. The power of the network is to open competition to non-traditional candidates.
  1. Scale networks
    This final category focuses on using networks to leverage local capabilities to address big problems more effectively.
    1. Replication networks make it possible to enable a concept or an innovative solution that has been successful on a small scale to spread rapidly. A dramatic example is TEDx, which took the model of an event that highlights “ideas worth spreading” from a limited (but highly successful) commercial venture to an “open” format that has been replicated internationally over ten thousand times.
    2. Coordination networks are designed to align and magnify the effectiveness of multiple ongoing efforts to address a common problem. A good model is the Global Alliance for Vaccination and Immunization (GAVI) that encourages more strategic responses to fighting infectious diseases globally by improving the quality of vaccines and making them more affordable.
    3. Cumulation networks take large tasks and divide them into smaller parts that many people can work on simultaneously. Amazon’s Mechanical Turk performs this function for relatively simple projects, while open source communities (such as the one supporting Linux) enable many developers to share responsibility for maintaining and enhancing valuable software programs.

Network rules. Slaughter’s typology categorizes networks in terms of positive goals, and there is no question that networks are being used for many constructive purposes. But networks can have unintended consequences or be intentionally used for anti-social purposes. And what is considered a worthy purpose in one environment (e.g., the promotion of democratic dialog) may be viewed as a threat in another environment.

As we have become more aware of the power of networks, and particularly of their darker side, we have entered into what Tom Wheeler described as “a race to make rules.” He also noted that an absence of rules does not mean “deregulation;” rather it means that “someone else is making the rules,” where someone else may be a commercial entity, like Facebook, or another government.

Historically, the Internet has been governed largely through voluntary groups such as the Internet Engineering Task Force (IETF) that focused mainly on setting standards. While this approach worked well as long as the Internet was largely U.S.-centric, it has become more fraught as the Internet has spread globally. Decision-making in these bodies has become politicized as countries like China attempt to influence the governance process of these supposedly independent bodies. And as the impact of the Internet grows, this voluntary approach to governance may be less effective for content or economic issues than for purely technical issues.

Because the Internet is now global, inaction by the United States may lead to action by others. A recent example is the fine imposed on Google by the European Union for unfairly favoring its own sites in search results. Another example is the broad personal privacy protection regime—including “the right to be forgotten”—imposed by the EU, which goes well beyond the limited protections provided by U.S. law.* Another type of initiative that threatens the integrity of the Internet is the emergence of data localization laws requiring that data generated in a country be stored in that country.

As Alec Ross, Distinguished Fellow at Johns Hopkins University, noted, norms vary greatly from country to county. For example, transparency and freedom of expression are important values in the U.S., but are much less so in many other countries. Protecting personal privacy is a higher priority in Europe than in the U.S. Both Facebook and Wikipedia are currently banned in China, while WhatsApp has been blocked in Brazil. Virtual private networks (VPNs), seen in the U.S. as indispensable tools for protecting privacy and legal in most countries, are forbidden in Russia, Belarus, China, Turkey, Iraq, United Arab Emirates and Oman. Facebook’s Monika Bickert commented that balancing the desire to operate under a single set of global standards with the need to conform to local requirement has become a formidable challenge for companies like hers.

Economics also play a large role in shaping regulations. Data localization rules are often motivated by a desire to foster a local IT industry, even when restricting the free movement of data run directly counter to the promise of the cloud to provide users with access to the most efficient resources available, wherever they happen to be located. And economic considerations are typically involved when powerful incumbents succeed in getting regulators to protect their interests from upstart providers with new technologies. This is particularly true when a government has a direct stake in one or more of the industry participants. Although liberalization has greatly reduced government ownership of telecom across the world, some governments retain a financial stake in their country’s network providers, which inevitably influences their approach to regulation.

Because these social, political and economic factors play out differently in different places, countries vary widely in how they regulate network activity. As a result, the Internet runs the risk of becoming balkanized. The question now, according to Esther Dyson, is whether the Internet will remain as “one beautiful global, but diverse, network” or whether it will devolve into many vertical, locally-controlled networks.

Shifting perspectives. As long as the Internet was perceived as a heterogeneous decentralized network that no one owned and no one controlled and was seen mainly as a playground for the digerati, it was generally viewed as a non-critical threat by most governments. But as the Internet became perhaps the most massive of all media and has come to be dominated by a few large, powerful and highly visible platform companies, the issue of control is more salient. Just as Willie Sutton robbed banks because that was where the money was, the Internet and its key platforms are important because that is where people’s attention is going. (A large majority of Americans regularly access the Internet, and the average user currently spends 4-5 hours per day online, mostly via mobile phones, and about half of that time on social media.) Concern about the power of networks has also grown as Internet-borne information warfare from both state and non-state actors has emerged as a serious threat over the past several years.

* Because so much Internet activity is international, the U.S. government found it necessary to negotiate a bi-lateral agreement, known as the Privacy Shield, with the European Union, in order to ensure compliance with EU privacy rules. U.S.-based companies that wish to transfer information from the EU to the U.S. can voluntarily subscribe to certify that their policies conform to EU law.

Share On