Political Economy & The State
Networks and Hierarchies

Has political hierarchy in the form of the state met its match in today’s networked world?

Appeared in: Volume 9, Number 6 | Published on: June 9, 2014
Niall Ferguson is Laurence A. Tisch Professor of History at Harvard and senior fellow at the Hoover Institution. His most recent book is The Great Degeneration: How Institutions Decay and Economies Die (Penguin Press).
show comments
  • Boritz

    “Nature means in this context the material or environmental constraints over which we
    still have little control, notably the laws of physics, the geography and geology of the planet, its climate and weather

    The NYT has huge level of control over climate and weather to the point where their faithful are not just concerned but alarmed.

    But speaking of the movie:
    You remember that scene where the robot Maria (actually an android with a realistic outer skin) is burned at the stake as a witch by the network types? She laughs maniacally as the flames
    consume her human looking skin revealing the metal body underneath and spoiling the ruse that she was the real Maria who leads the rebellion. That was just awesome.

    The film predicts humanoid robots by 2026 as well as the burning of witches a hundred years into the future of the film producers.

    You can have in any combination:
    a good hierarchy
    a bad hierarchy
    a good network
    a bad network

    The interplay between the two is worthy of study as Ferguson has done and the “quiet accommodation” may be the best that can be hoped for to keep both semi-pseudo-honest.

    • Andrew Allison

      Hierarchies and networks are organizational structures. It is the uses to which they are put which are good or bad. Unhappily, human nature being what it is, Metropolis does indeed live on.

  • Anthony

    There is “history’s fundamental dialectic between hierarchies and networks” to which I offer of course metropolis lives on – as their is always present a politico-economuc system or hierarchies with economic ramifications and economics (networks) with political ramifications (as Ferguson intimately alludes). To this end and sustaining its perpetuation remains reality, people usually are too inattentive, unperceptive or lacking in judgment to be able to distinguish potential friend (netizens) from betrayers (hierarchies) – “the narcissism of small differences”. Which underscores “how can an urbanized, technologically advanced society avoid disaster when its social consequences are profoundly anti-egalitarian”. That is unless said society becomes more attentive (mindful) so that our social behavior as citizens (netizens) reflect organizing approach to our lives and socio-economic arrangements positively challenging what’s described as anti-egalitarian outcomes, some negatives social consequences are unavoidable. As an illustration, the networks (perhaps in combination with hierarchies) “cater to our solipsism (selfies), our short attention spans (140 characters), and our seemingly insatiable appetite for news about celebrities” thereby maintaining a distracted society. In short, the technologies of mass persuasion or the enticement of modern media serves whom? The question what gets attention becomes neural rewiring to what end (“we are not just what we eat. We are also what we see and hear since these literally change our brains, minds, and future judgment.”).

  • Mr. Ferguson is misinterpreting the key properties of networks and hierarchies. That in turn is producing an historical interpretation that’s not as useful as it might be.

    First, hierarchies are networks. What needs to be distinguished between the two is whether or not the network is scale-free. Hierarchies are not scale-free; they’re engineered by humans to achieve specific, human-directed tasks. Scale-free networks (the internet, social networks, free market economies, the relations between nation-states, ecosystems, etc.) evolve, and the tasks that they accomplish are unplanned. Scale-free networks are vastly more efficient at performing a “task”, but the downside is that you don’t really get to pick what the task is.

    If you want to run a government or fight a war, you want a hierarchy. You can’t direct a scale-free network to do something. You might want to fight Nazi Germany, and a scale-free network might evolve to purchase leasing rights for oil in Indonesia instead. Not only do hierarchies work better at accomplishing specific tasks, they provide an engineered framework in which the “owners” of the hierarchy can deploy huge resources (either through recruiting motivated volunteers or through coercion) and understand what they’re doing.

    It’s also important to understand that hierarchies are often best understood as individual nodes in a larger scale free network. The US and Chinese governments may be hierarchical, but the nations themselves are merely nodes in a network of other nation-states, which in turn is augmented with corporations, NGOs, interest groups, terrorists, social networks, scientific societies, and a host of other actors.

    Seen from that light, the proper question isn’t whether Ferguson’s “networks” are going to overwhelm his “hierarchies”; it’s how important the hierarchical nodes in the larger scale-free network are in relation to other actors in the network.

    Where this gets really interesting is when you start to consider the fundamental property of scale-free networks: the distribution of the nodes when ranked by the number of connections they have to other nodes in the networks. In scale-free networks, this distribution always follows a power law, which means that a very small number of nodes are massively connected, and there’s a long tail of less-connected nodes that falls off very rapidly, with the vast majority of nodes having a very small number of nodes.

    In the Great Scale-Free Network of Life, “power” is roughly proportional to “number of connections”. When viewed this way, Ferguson’s argument can be re-cast into a simpler form: Hierarchies, with their ability to enlist vast resources for specific purposes used to be (and still are) the most powerful, most connected nodes in the network. But technology has allowed other entities to achieve very high connectivity very quickly, allowing them to compete on near-equal terms with the hierarchically organized entities.

    In a perfect world, we’d learn how to engineer scale-free networks for specific tasks, which would result in the replacement of the current hierarchies with systems that were more efficient and adaptive than our current governments and corporations. But we don’t know how to do that yet. Until we do, we’re likely to have to live with the inefficiencies of hierarchy. However, these inefficiencies should be encouraging us to make these hierarchies more and more focused on doing as limited a set of tasks as possible, as well as possible, rather than trying to create hierarchies that pervade every aspect of our lives and provide mediocre results. Scale-free networks may not be able to be directed to do good things, but their efficiencies and adaptability tend to make a lot more good things happen than bad. If you could make that statement about most of our hierarchies, that would be a big step forward.

    • Curious Mayhem

      Great answer. The typology of networks is richer than intimated here. There’s the random network, the “small world” network, the pathfinder network, and so on.

      Hierarchies are indeed a type of network, optimized to achieve particular ends. They are not optimized in a general sense, for flexibility or longevity. The hierarchical systems of the 20th century were, initially, optimized for one thing, to fight and win “total” wars (and here I include “total social revolutions” as a kind of war, as “people’s revolutions” and “people’s wars” were inseparable). Later, they were secondarily optimized to achieve the necessary level of social integration to those primary ends. They reached the zenith of their legitimacy and popularity in the 1930s and 40s and the peak of their power and reach in the 1960s.

      While they still have a lot of power, hierarchies have lost much of their legitimacy, while also frittering away much of their power.

      • Thirdsyphon

        Hierarchies might not be optimized in a general sense for flexibility, but (at least in environments where flexibility isn’t a primary requirement for survival), they are optimized for longevity. See, e.g., the Roman Catholic Church.
        Successful human organizations across history have been hierarchical more often than not. Some form of feudalism seems to spontaneously emerge wherever and whenever sufficient resources become available to support it.

    • Thirdsyphon

      “If you want to run a government or fight a war, you want a hierarchy. You can’t direct a scale-free network to do something. You might want to fight Nazi Germany, and a scale-free network might evolve to purchase leasing rights for oil in Indonesia instead.”
      Or, since the Indonesian oil reserves of the period were firmly under the control of Imperial Japan, a scale-free network thirsty for oil might proritize waging war against Japanese imperialists over waging war against the Nazis. . .which is exactly what the United States wound up doing.
      Maybe the real question here is how much “non-scale” network decisionmaking creeps into and affects what are ostensibly “scale” network decisions. . . i.e.: who’s really in charge? In human governments, as in human neurology, the answer might be more complex and far stranger than we think.

      • Actually, the “Europe first” agreement held up very well, even though it was arguably not in America’s immediate national interest.

        “Scale-free decision-making” is to some extent an oxymoron. The beauty of network effects is that individual decisions by smart actors get accommodated in strange and wonderful ways by the network, which leads to emergent (and usually beneficial) properties that no single actor would ever have thought of. Hierarchies don’t do that.

        That’s not to say that hierarchies don’t generate good ideas and act on them. But the mechanism is completely different. Hierarchies innovate when somebody has an idea, feeds it up to the proper level, and a decision-maker decides to reallocate resources in the hierarchy based on the new idea. Scale-free networks adjust to innovations by changing the connections in response to the actor with the new idea. That often means copying the innovation and modifying it slightly, but it can also mean forging new relationships that leverage the innovation for unanticipated purposes. In either case, there’s no “decision-maker”.

        I want to be careful not to denigrate the use of human intelligence in hierarchies. A smart decision-maker can sometimes accomplish things faster and more effectively in a hierarchy. Some of those things simply aren’t possible to accomplish in a scale-free network. The invention of hierarchical organization has been–and continues to be–a huge boon to human civilization. But hierarchies are only good for solving problems when you know what the problem is. Scale-free networks tend to generate answers much slower than hierarchies, but they’re really, really good at identifying the right problems.

    • Amrendra Kumar

      Dear RadicalModerate, thanks for the insightful comment. Your point that hierarchies exist within free-scale and that hierarchies exist to execute to a stated goal makes more sense than Niall’s argument.

    • jnk9

      In scale-free networks, this distribution always follows a power law, which means that a very small number of nodes are massively connected …

      This is a most important point, and Ferguson touches upon it when noticing the concentration of power in the hands of Google and Facebook.

      Members of a network have no trouble identifying its “leaders”, and a comparison between the leaders of a network and the commanders of a hierarchy resembles the scene in Animal Farm when the animals look through the window at the pigs and the farmers, and cannot tell one from the other.

      Early adoption of a novel mechanism for networking is a most effective way to create a new hierarchy with oneself at or near the top. However, the leadership position is fragile: Google and Facebook displaced Altavista and MySpace, and may yet be displaced themselves.

      • jnk9

        Conspiracy theories often propose that society is controlled by the members of a small secret organization.

        If you think you see evidence for this theory, often you are simply seeing the power-law distribution of connections in a scale-free network. One such “secret organization” is hidden in plain sight: it comprises the powerful and well-connected people who socialize together, usually in or near a country’s capital city.

        Of course, sometimes the secret organization does exist – for example, “P2” in Italy.

      • “Early adoption of a novel mechanism for networking is a most effective
        way to create a new hierarchy with oneself at or near the top.”

        There’s a subtle distinction that needs to be made here. While Google and Facebook are highly-connected nodes, their power doesn’t magically allow them to start organizing chunks of the network into a hierarchy for them to command. They may, through lobbying, leverage on their developers, and use of their user bases, be able to construct a hierarchy (or at least something that is scaled), but that power only exists to the extent that the scale-free network in which they’re embedded finds them useful and provides them high connectivity. Remove the connectivity and the commanded portions of the network will dry up and blow away–as they should.

        • jnk9

          An interesting distinction. However, the capacity to command, although limited, is real. Highly-connected nodes such as Google and Facebook can earn an income stream from their position, and can organize a chunk of the network to a limited extent: for example in the range of their services, the terms and conditions, and the character of the advertising. Obviously, if they become too coercive, their customers (less well-connected nodes) will go somewhere else. The Achilles heel of both companies is their use of personal information, and the intrusiveness of their advertising – serving personalised advertising that tracks a user around the Web. I suggest that Google has struck a better balance than Facebook in this respect, probably because it has longer experience, and a better understanding of networks. Facebook is still learning, and is more likely to be usurped by an innovating newcomer.

          Another point that occurred to me is the similarity between advertising revenue, and the apparently benign system of micropayments for content proposed many years ago by Ted Nelson. In practice the latter would have to have aggressive controls to prevent “click fraud” – principally from botnets – and these would be hard to distinguish from the measures deployed by Google. Perhaps advertising is the only viable system of micropayments, because the suppression of click fraud is strongly motivated by the defence of the advertising company’s income stream and share price.

  • Reza Mahani

    Amazing piece full of insight. I may not agree with all points (e.g. we may actually want a balance between networks and hierarchies rather than network dominance) but there are many valuable points that are woven together in a coherent narrative. Thanks much!

  • Amrendra Kumar

    Niall Fergusson never disappoints, Network Vs Hierarchy is very apt conceptual framework to look at impact of technology on Society.

  • attagirlny

    Too long to read.

  • johnwerneken

    The only weak spot are the finances: debt and deficit are unsustainable, at least in a rights and vetoes and no growth for all economy. The currency would be worthless, if there were any credible alternative. The crypto currencies MIGHT become an effective payment mechanism; with sufficient use, perhaps a store of value and a re-birth of a true currency after 100 years of B.S. instead of money. The rest of it is just the organizational challenge writ large, which the cleverest hierarchs have always mastered quite nicely: domesticate autonomy. Not for nothing are the citizens often referred to as the American Sheeple.

  • John Lofton

    But there is one “hierarchy” we must not forget — the Triune God. We must not forget that the Christ is Lord, King of kings, Lord of lords, the One Who has ALL power, in Heaven and on earth.

    John Lofton, Recovering Republican
    Dir., The God And Government Project
    [email protected]


  • mitspanner

    The State as a hierarchical institution is best at accreting power and crushing all resistance to that end. All of its other activities are subordinate to that. The increasing networking capability of individuals is coming to a point where it can rival the State in terms of resisting it or avoiding it. Hierarchy will not be eradicated, rather it will be subject to more and more competition from market forces. I am more positive in my outlook on this score than Mr. Ferguson.

  • Hierarchy and network are very much connected, in the sense of a primordial family , where networking exists on everything, yet decisions might be delegated to some rather than all members on certain or selected topics of daily life. The scale free concept enriches the argument, however in my perception there is a continuous interplay , between networks of all kinds and the hierarchies they tend to interact with. A person watching television is part of a network of audience, a team playing soccer is part of a show, an administration expressing governance is part of an iceberg in hierarchy that interacts with different networks, local, national , multinational. Evolving the interplay might be affected by innovation, technology or science, however the basic is that individuals tend to socialize with peers, create dynamic networks on determined options. These options/ goals are pursued by a network/hierarchy ad-hoc that is firmly or less firmly guided by upper levels of hierarchy exerted either by practicality at first or by exploitation. Indeed a very nice article. Thank you.

© The American Interest LLC 2005-2018 About Us Masthead Submissions Advertise Customer Service
We are a participant in the Amazon Services LLC Associates Program, an affiliate advertising program designed to provide a means for us to earn fees by linking to Amazon.com and affiliated sites.