Thoughts on Wu, Part 5: What Ultimately Separates the Cyber-Libertarian & Cyber-Collectivist

by on October 29, 2010 · 15 comments

I want to thank Tim Wu for continuing to engage in a discussion here about his book, The Master Switch, with his various comments to my ongoing rants.  After pouring out about 15,000 words over the past 4 days, I suspect I’m beginning to sound a bit like his cyber-stalker!  I feel a bit bad about this because I really do like Tim a lot and find him to be one of the all-around coolest and most laid-back guys in the Net policy business.  But, as I’ve noted in my ongoing series [see parts 1, 2, 3, & 4], we have profoundly different worldviews when it comes to information history and policy. And some of the recent comments he made to my 3rd post deserve a serious response.

In one of those comments he asks, “The question, then, is how you get, essentially, limited, controlled government in regulatory affairs; how you duplicate, in some sense, the limits imposed on other dangerous gov’t functions like the army. I don’t think this is having things both ways; I think this is trying to learn from what has gone wrong in the past.”  In the other, he says: “The question I’m asking in the end of the book is whether we can do better; try to have rules against the worse forms abuse without a creeping regulation that turns into capture. I suspect you think that’s impossible, but I don’t.”

So, here’s my response (and I’m making it a new, dedicated post here instead of just a comment in an old thread because I feel we are getting to the heart of the difference between cyber-libertarians (like myself) and cyber-collectivists (or whatever Tim would call himself).

To be clear, I don’t think corporations are angels or that there is never a time when a market can’t be naturally subject to a great deal of control by one company or a handful of companies.  The difference between us comes down to two things primarily.

First, as I have already noted in a couple of these essays (especially this one), I believe regulatory capture, mismanagement, or other shenanigans have more to do with creating and / or maintaining “monopoly” or lasting / harmful “market power” than natural market forces.   By definition, a “purely economic laissez-faire approach” does not exist in markets characterized by regulatory capture and bureaucratic mismanagement.  And you won’t ever get less regulatory capture and bureaucratic mismanagement by increasing the scope of government control over a market.

Second, to the extent that any company or set of companies is able to achieve “market power” is a largely natural fashion (think IBM in 70s or Microsoft in late 90s), I believe that markets can and do act to evolve around those situations quite rapidly, even more rapidly when the market is built on code.

I spent time developing these points in detail in this two-part debate [1, 2] with Lawrence Lessig, which I hope Prof. Wu will take the time to read since I went to great pains to clearly delineate the differences that separate our worldviews.  Ultimately, as I said there in response to Prof. Lessig, what really separates the cyber-libertarian and cyber-collectivist schools of thinking comes down to a belief that “market failures” or “code failures” are ultimately better addressed by voluntary, spontaneous, bottom-up, marketplace responses than by coerced, top-down, governmental solutions. Moreover, the decisive advantage of the market-driven approach to correcting code failure comes down to the rapidity and nimbleness of those response(s).

Does that mean cyber-libertarians believe everything will be all wine and roses in a truly free marketplace?  Absolutely not.  There will be short term spells of what many of us would regard as excessive market power.  The difference between us comes down to the amount of faith we would place in government actors versus market forces / evolution to better solve that problem.  We cyber-libertarians would obviously have a lot more patience with markets and technological change, and would be willing to wait and see how things work out.  We believe, as I have noted in my previous responses to Wu, that it is during what some regard as a market’s darkest hour when some of the most exciting disruptive technologies and innovation are developing.   We are bullish on what I have called experimental, evolutionary dynamism.  People don’t sit still; they respond to incentives, including short-term spells of “market power.”

Is this blind faith in the market?  I suspect Prof. Wu and others would accuse us of that.  But I would argue it isn’t blind faith but informed fact.  It’s interesting, for example, that one of the “information empires” Wu doesn’t spend much time on in his book is IBM.  Back in the 60s and 70s, (as I have documented here before) IBM was the big, bad dog of the computing world, with significant “market power” in mainframes — the only computers that really counted at the time.  Big Blue’s market power was achieved in a fairly natural way, however.  Importantly, there isn’t much regulatory capture or interference I could point to that helped cause or maintain the power IBM had. So, it’s certainly a better case study than others Wu uses in his book, most of which were subject to early meddling by government that tipped the balance in unnatural directions.

Anyway, back in the 1960’s, some folks at the time feared IBM might “leverage” their significant market power into new fields. As a result, the Department of Justice opened an antitrust case against Big Blue in 1969 that would become a 13-year quagmire, with little to show for all the legal wrangling by the time the case was abandoned in 1982.  Here’s how CNet staff writer Rachel Konrad summarized the fiasco back in 2000:

In January 1969, the government began a sweeping antitrust investigation into IBM’s dominance and attempted to break it into smaller companies that would compete against one another. During the six most critical years of the trial, from 1975 to 1980, the parties called 974 witnesses and read 104,400 pages of transcripts, according to Emerson Pugh’s 1995 book “Building IBM: Shaping an Industry and Its Technology.”

The 13-year investigation, which required IBM to retain 200 attorneys at one point, fizzled in the early ’80s as the computing landscape shifted from mainframes to personal computers. The government abandoned the tainted effort entirely in 1982, as clones of the IBM PC eroded Big Blue’s dominance. But the company, still fearful of the watchful eye of the Justice Department, took pains to avoid the appearance of a monopoly long after it relinquished its hold on the market. People who worked for IBM in the ’80s and early ’90s said the company routinely fell victim to “pricing death strategy”–a reluctance to lower prices below cost, even on products that weren’t selling–to avoid what the government would call predatory pricing. By the mid-’80s, the company was in bad shape. The antitrust troubles, combined with ill-timed product failures such as the Future System, pinched revenues. The company began a nearly decade-long financial slide. In retrospect, the antitrust case against IBM seemed laughable.

IBM had become the victim of a classic “disruptive technology” paradigm shift that few could have foreseen in 1969.  As Peter Pitsch noted in his 1996 PFF book The Innovation Age, “In 1981 the Department of Justice was still pressing their case against IBM while market forces were about to lay waste to the company.” Pitsch continued:

IBM certainly did not expect to see PCs erode the market share and profitability of its venerable mainframe computers, but the fall of the old “big iron” machines was rapid and spectacular. The revenue of IBM’s mainframe unit fell from roughly $9 billion in 1990 to an estimated $4.5 billion in 1994… [T]he parties destined to become players in the PC revolution were unknown when the PC was introduced, and the experts’ predictions of a much-ballyhooed computer face-off between IBM and AT&T never materialized. Innovative companies that did not exist at the beginning of the revolution rose rapidly. Few people had ever heard of a small company named Microsoft. Nor had they heard of Intel, Novell, Compaq, Dell, or Netscape.

Pitsch went on to summarize how IBM’s manufacturing capacity was slashed in the years that followed and also notes that, astonishingly, “in the space of five years after 1987, IBM lost two thirds of its market value — more than $70 billion.”  In sum, new marketplace innovation and competition handled the short-term market power concern that antitrust regulators had about Big Blue.  Pitsch goes on to explain what the antitrust regulators missed:

A dominant firm can lose its “King of the Hill” status in two ways. First, if it does not continually improve, it will lose market share and profits to low-cost imitators. For example, the ability of low-end PC manufacturers to make IBM clones fostered robust price competition in the PC market. Second, today’s market leaders must worry that some established and well-financed competitor or possibly an upstart produce a technical breakthrough that will displace them. This situation reflects [the] fact that gains from innovation are so powerful and beneficial to consumers that they outweigh the higher prices dominant firms can charge. Indeed, attempts to eliminate these high profits by regulating prices would almost certainly disserve consumers even if the regulations dampened the incentives for innovation only slightly.

What Pitsch is talking about here is dynamic competition, not the static competition. And what the history of IBM shows is the power of evolutionary dynamism in action.  Markets are a learning experience; a “discovery process” as Austrian economists have taught us. Those of us who believe in dynamic competition and evolutionary dynamism see markets in a constant state of flux and expect that sub-optimal market developments or configurations are exactly the spark that incentivizes new form of market entry, innovation, price competition, and so on. Experimentation and evolution happen if you let them happen.

Others, however – and I suspect this includes Prof. Wu – would argue that’s not good enough. They want action, and they want it now!  Every short-term hiccup deserves a policy response in the name of protecting “the public interest,” however they define it through regulation.  But what about the costs and trade-offs associated with early, preemptive, or prescriptive regulation?  What of the danger of regulation steering markets in unnatural or inefficient directions? The possibility of picking technological winners and losers, or technological lock-in?  The possibility of regulatory capture and the creation of a special interest, lobbying hell inside the Beltway?

Somehow these factors often go out the window for those who subscribe to the more static, snapshot-oriented view of markets and competition that is so prevalent in cyber-collectivist circles.  But the cyber-libertarian can’t let those go.  Those factors lie at the core of the problem, we would argue. Actions have consequences. Regulations have costs. And those costs typically outweigh the benefits of preemptive strikes by the State.

And that, at root, is what separates the cyber-libertarian and cyber-collectivist worldviews when it comes to concerns about “market power” and what to do about it.

_____________

[Jump to Part 6 in the series.]

  • Concerned

    Adam,

    Are you familiar with the book Cyberselfish, written back in 2000 as a critique of cyber libertarianism and the whole Silicon Valley culture? In case you are not, definitely check it out – you are the living stereotype the author criticizes on the book. Your anti-regulation bias is so strong you do not even see the incredible contradictions in your own argument. Besides, you seem to ignore completely how other countries have approached Internet regulation and how it has worked out for them.

    Do not get me wrong – you do bring up some interesting points and counterarguments. But the whole anti-regulation rant – while quite the real for the existence of this blog – is simply ignored on the international level.

  • http://bennett.com/blog Richard Bennett

    It's not clear that other countries that have pursued a more restrictive regulatory approach than the US has in networked markets have actually benefited at all. In fact, most market-based economies regulate Internet service pretty much the same way the US does, including the use of some ideas that were developed in the US and then discarded. I assume you don't hold Iran up as a model, even though their close scrutiny and control of the Internet follows the model you seem to endorse, “Concerned.” The unbundling model in Europe has failed to produce high-speed Internet access, for example, and for that reason some European countries are pursuing a publicly-financed “dark fiber” model in their big cities.

    The crux of the problem is the issue that Tim and Adam are debating: how do you keep your government regulator from corrupting and impeding progress while constructively accelerating technology in advance of the markets than can sustain it?

    It's not an easy question.

  • larrydownes

    Amazing work, Adam, and some essential points raised in the course of your reviews. (It's even longer than a New York Review of Books review, which are endless!)

    As a consultant with Accenture throughout the 1980's, I worked exclusively with IBM equipment in developing and installing systems for clients. The hangover of the antitrust investigation, and, indeed, the 1952 consent decree, were an everyday experience. The divisions had strict Chinese walls between them. I worked mostly with GSD, the mid-range group. On the expectation that the DoJ was going to break them up by division, GSD had invested heavily in what they called the Future Systems project–a new line of computers that would have features that would still be considered state-of-the-art, including built-in relational database, object-level security, and high-level interface to the operating system.

    When DoJ dropped the investigation, GSD was left holding the bag, and DPD, fearing cannibalization, refused to let them exploit any of the technologies that had been invented beyond a crippled, small computing system known as the System/38.

    There's a great irony to the fact that being broken up, AT&T thrived. Left together, IBM worried over a future break-up so badly that they did more damage to themselves than any competitor could have.

  • http://bennett.com/blog Richard Bennett

    The System/38 was a wonderful little computer, well ahead of its time and much too advanced for its scale. System architects at other firms studied it like mad and tried to apply the capability architecture where they could.

  • Larry

    Ah, Richard! I knew there was something about you I really liked! I still have parts of the CPF Programmer's Guide memorized.

  • Tim Wu

    Thanks for this post, which I do agree gets to the crux of the problem.

    It is a truism that all things change. In economics, every cartel will always, eventually, fall apart. Every monopoly will lose its market; every once dominant firm will eventually be in the dustbin. Adam's post suggests the merits of patience.

    I take this point; it is a good one. Regulators in the past were definitely to eager to try and solve problems that might have solved themselves.

    That does leave two hard questions. First, whether it is always worth waiting out the cycle. Damage can be done in the interim, while waiting for the cartel to fall; everyone is paying higher prices. And if a new cartel forms, and a new one — eventually, it can be more cartel than competition.

    Second, when a monopolist has seized control of regulatory power to protect itself, there is no way anyone can expect the market to displace it. The firm has then effectively made itself part of government.

    In such cases usually only Government can destroy what it itself created. This is what my support for the AT&T breakup in the book is premised. At some point it is no longer an issue of market versus government – it is government restraining its own creature — something we also call “separation of powers.”

  • http://bennett.com/blog Richard Bennett

    It's interesting that the thing that ultimately broke IBM's computing monopoly was the IBM PC. The competition for fame, glory, and promotions within the company managed to produce a platform-buster with an open architecture. IBM's decision to publish the source code of the BIOS was crucial to the ability of other firms to produce applications, but it also enabled Compaq and others to produce PC clones. Similarly, the specs for the expansion card bus were made public, which enable third parties like Hercules to create an aftermarket in addon cards. The parallel in the telecom space is Carterfone.

    In technology markets, altering the organization of the monopoly firm may not be as important as publishing interfaces that allow for others to nibble away at its technology.

  • Jim Harper

    A few comments on this reply, Professor Wu, which I find rather thin.

    You say that regulators “in the past” were too eager to jump in, from which I infer that you think they’re not too eager today. Is there any way of making that an empirical claim? Or is it simply your opinion that their interest in regulating these days is merited where it wasn’t before?

    There are only a few ways you can hold the latter view. 1) You are omniscient and so you know that market responses to current conditions would, if allowed, produce lower overall welfare than regulatory responses. (This is unlikely – no offense.) 2) You ignore the lessons of history, simply imagining a present and future where regulators respond differently to the same incentives. 3) You have identified a new set of conditions in which the behavior predicted by Niskanen’s Bureaucracy and Public Economics no longer occurs.

    Which is it? And it would be great if you could explain why.

    As to your questions, it would be helpful for you to identify, categorically or by example, what “damage” you think can occur if a cartel is not undercut by government. The natural influence of cartels, as Adam wrote about, is to invite competition — the higher its prices, the stronger the invitation — and substitutes — people moving to alternative goods and services. If the government attacks an emergent cartel and prices never rise, new competitors don’t enter the field and consumers don’t identify substitutes. Accordingly, new, lower-equilibrium prices are never reached. Consumers are worse off than they would have been under a short period of cartelization or monopoly. That’s the “damage” produced by government intervention—unseen but profound.

    In your second “question”—I’ll confess I don’t know what the actual question is—I sense you giving independent motive to government, as if government is not a product of all the influences brought to bear on it. Are you a believer in “scientific government,” with cadres of independent regulators using the best science has to offer free of political influence? This Progressive-era view of regulation has gone by the wayside, debunked by, among others, Stephen Breyer, writing before he went to the Supreme Court.

    I’ll bet that any instance of government “destroy[ing] what it itself created” could—with enough time and research—be revealed as competitors and potential competitors finally dismantling the regulatory apparatus that a monopolist had erected around itself. I think of MCI’s wily effort to get FCC to approve what ended up being competition for AT&T’s long-distance telephone service. My question is: Why create the conditions for competitors to use government power, requiring potential competitors to devote decades of work and countless millions to get a chance to improve life for consumers?

    I don’t know your thinking well enough to say that this critique applies, but I feel like it might: http://tinyurl.com/29jce4d

    What this has to do with separation of powers, I do not know. All three branches were involved (in ways) in the break-up of AT&T, but that history is nothing like the contest for power the Framers set up with our tripartite government. Unless I’m mistaken the demise of the ICC and the CAB were mostly executive-branch driven, and did not result (much) from inter-branch rivalry. Does not compute.

    Adam’s done yeoman work to critique your book. (Do understand, he’s always a little bombastic—except when it comes to praising fellow TLFers.) I think he should get a full response, either here—I’m sure we’d give you a guest post—or in the venue of your choice.

  • Jim Harper

    A few comments on this reply, Professor Wu, which I find rather thin.

    You say that regulators “in the past” were too eager to jump in, from which I infer that you think they're not too eager today. Is there any way of making that an empirical claim? Or is it simply your opinion that their interest in regulating these days is merited where it wasn't before?

    There are only a few ways you can hold the latter view. 1) You are omniscient and so you know that market responses to current conditions would, if allowed, produce lower overall welfare than regulatory responses. (This is unlikely – no offense.) 2) You ignore the lessons of history, simply imagining a present and future where regulators respond differently to the same incentives. 3) You have identified a new set of conditions in which the behavior predicted by Niskanen's Bureaucracy and Public Economics no longer occurs.

    Which is it? And it would be great if you could explain why.

    As to your questions, it would be helpful for you to identify, categorically or by example, what “damage” you think can occur if a cartel is not undercut by government. The natural influence of cartels, as Adam wrote about, is to invite competition — the higher its prices, the stronger the invitation — and substitutes — people moving to alternative goods and services. If the government attacks an emergent cartel and prices never rise, new competitors don't enter the field and consumers don't identify substitutes. Accordingly, new, lower-equilibrium prices are never reached. Consumers are worse off than they would have been under a short period of cartelization or monopoly. That's the “damage” produced by government intervention—unseen but profound.

    Markets move forward the way people walk—by instituting a forward fall and then throwing a leg forward before the face hits the ground. If falling forward is preempted, steps are smaller, or they don't happen at all. That's safe, but not progressive.

    In your second “question”—I'll confess I don't know what the actual question is—I sense you giving independent motive to government, as if government is not a product of all the influences brought to bear on it. Are you a believer in “scientific government,” with cadres of independent regulators using the best science has to offer free of political influence? This Progressive-era view of regulation has gone by the wayside, debunked by, among others, Stephen Breyer, writing before he went to the Supreme Court.

    I'll bet that any instance of government “destroy[ing] what it itself created” could—with enough time and research—be revealed as competitors and potential competitors finally dismantling the regulatory apparatus that a monopolist had erected around itself. I think of MCI's wily effort to get FCC to approve what ended up being competition for AT&T's long-distance telephone service. My question is: Why create the conditions for competitors to use government power, requiring potential competitors to devote decades of work and countless millions to get a chance to improve life for consumers?

    I don't know your thinking well enough to say that this crtique applies to your thinking, but I feel like it might: http://tinyurl.com/29jce4d

    What this has to do with separation of powers, I do not know. All three branches were involved (in ways) in the break-up of AT&T, but that history is nothing like the contest for power the Framers set up with our tripartite government. Unless I'm mistaken the demise of the ICC and the CAB were mostly executive-branch driven, and did not result (much) from inter-branch rivalry. Does not compute.

    Adam's done yeoman work to critique your book. (Do understand, he's always a little bombastic—except when it comes to praising fellow TLFers.) I think he should get a full response, either here—I'm sure we'd give you a guest post—or in the venue of your choice.

  • Pingback: Thoughts on Tim Wu’s Master Switch, Part 4 (on Regulatory Capture)

  • Pingback: Thoughts on Wu’s Master Switch, Part 6 (His Audacious Information Industrial Policy)

  • Pingback: An Invitation to Tim Wu to Respond

  • Pingback: Tim Wu Redefines “Monopoly”

  • Pingback: How Should Libertarians Think about The Master Switch?

  • Pingback: Some Random Thoughts on AT&T / T-Mobile Merger

Previous post:

Next post: