Organizations and Markets

10 Comments

• 1. Richard Makadok  |  24 August 2006 at 9:41 am

  Hi, Lasse. Thanks very much for your kind words about my working paper, and for helping to spread the buzz.

  You make a good point about entry barriers being a necessary condition for creating profit, and I believe that would be true under any of the four mechanisms I’ve identified — not just collusion.

  But on the other hand, entry barriers are definitely not a sufficient condition for profit under competition. A Bertrand duopoly with an undifferentiated product produced by symmetric profit-maximizing firms generates no profit for either player, even if it’s protected by perfectly impenetrable entry barriers. In order for those entry barriers to contribute toward the creation of profit under competition, one would have to change at least one of those adjectives — Bertrand, undifferentiated, symmetric, or profit-maximizing. Changing the “Bertrand” adjective implies a collusion-based mechanism. Changing the “undifferentiated” adjective implies what I call a competence-based mechanism. Changing the “symmetric” adjective implies either a competence-based mechanism (e.g., cost differentials) or a committment-timing-based mechanism (e.g., early-mover or late-mover advantage). Changing the “profit-maximizing” adjective implies a governance-based mechanism.

  Note that I use the phrase “profit under competition” in the manuscript, rather than just “profit.” If entry barriers are strong enough to prevent any competition at all — i.e., to create a true monopoly — then, yes, they really are a sufficient condition for profit (unless the monopolist suffers from a truly egregious governance problem that causes an enormous deviation from profit-maximizing behavior). But if the entry barriers admit any competition at all, then they no longer constitute, by themselves, a sufficient condition for profit. I am not interested in the case where entry barriers generate a true monopoly, because I’m interested in explaining profit under competition — not profit in the absence of competition. Explaining profit in the absence of competition is trivial — something that economists had figured out hundreds of years ago.

  So, entry barriers are a necessary but not sufficient condition for profit under competition. In order to generate an actual profit under competition, the entry barriers must be combined with one or more of these other four mechanisms. For this reason, I prefer to treat entry barriers as simply a “background assumption” rather than as a separate mechanism.

  Let’s take an analogy. If we were constructing a taxonomy of mechanisms for generating lift on fixed-wing heavier-than-air aircraft, we would have to agree that the presence of air around the aircraft is a necessary condition. Without air to flow through (e.g., in a vacuum), the wings can never produce any lift. But the presence of air does not, by itself, guarantee that the aircraft will actually lift off of the runway. In order for that to happen, you need some mechanism to propel the aircraft forward with sufficient thrust. To date, four mechanisms have been developed to do that — piston-driven propellor, jet turbine, rocket, and ramjet. Yes, we have to treat the presence of air as a background assumption, but the presence of air does not, by itself, get the aircraft off of the ground and therefore does not qualify for membership in a taxonomy of lift-generating mechanisms for heavier-than-air aircraft.

  In case it’s not perfectly transparent, in my analogy, air corresponds to entry barriers, and the four thrust-generating mechanisms correspond to the four profit-generating mechanisms.

  In this analogy, the only exception would be for the case of lighter-than-air aircraft like blimps, zeppelins, and balloons — where the mere presence of air is sufficient to generate lift. But note that I phrased it as “a taxonomy of mechanisms for generating lift on fixed-wing heavier-than-air aircraft” in order to exclude such lighter-than-air aircraft. But explaining lift in lighter-than-air aircraft is trivial — something that Archimedes had figured out thousands of years ago. In my analogy, lighter-than-air craft corresponds to monopoly — both being trivial, uninteresting cases.

  I do admit that I need to spell out all of this logic more clearly and explicitly in the manuscript, probably in the introduction.

  I hope this helps to clarify the logic.

  Thanks again,
  Rich Makadok

• 2. Peter Klein  |  24 August 2006 at 10:49 am

  Rich:

  First, let me add my own appreciation for the paper and the important and provocative issues it raises. Great work.

  Second, is there a version of the paper we could link to, for readers who haven’t yet had a chance to see it?

  Third, I’d like to push you a bit more on the classification exercise. As Lasse emphasized, the four theories are explanations for positive profits under equilibrium. As you mention above, some form of competition — not perfect competition, obviously, under which entry barriers are assumed away, but not pure monopoly either — is assumed, to make things interesting. Disequilibrium profits, of the kind emphasized by Knight, would then be in a fifth category. (Such profits can’t exist under perfect competition, but they don’t require entry barriers, unless we classify asymmetric information, or differences in “judgment,” to use Knight’s word, as an entry barrier.) If the object is to explain sustained competitive advantage, I can see why this category might be less interesting, though (a) its exclusion has seriously misled policymakers (Littlechild, 1981) and (b) it may be more important for strategy than most of us realize (Matthews, 2006).

  Is it appropriate to distinguish equilibrium profits from equilibrium rents, do you think? All equilibrium profits are monopoly profits, in the sense that sellers can somehow raise prices above marginal costs, regardless of how many firms are in the industry (your category 1). The same isn’t true of rents, or returns to superior factors of production (your category 2 and — if the gains from choosing superior governance structures are returns to superior management — your category 4 as well).

  In short, business income (accounting profit) can be decomposed into disequilibrium profit, monopoly profit, the business owner’s implicit wage, and (possibly) interest. Factor payments are rents, and are not included in accounting profits (other than the owner’s own implicit wage). Your exercise is to explain the interactions among one set of theories to explain monopoly profit (1) and one or two sets to explain factor payments (2 and maybe 4). (I’m not sure how to classify 3 in my scheme).

  Have I got this right?

• 3. Richard Makadok  |  24 August 2006 at 10:56 am

  One more thought…

  My only complaint about Lasse’s posting is that it largely overlooks the main point of the manuscript (mentioning it only very briefly, in passing, near the end of the posting).

  The main point of the manuscript is NOT the taxonomy of profit-creation mechanisms per se.

  Rather, the main point is that, across four decades of research, we have focused fairly exclusively on the “main effects” of these four mechanisms, while almost completely ignoring the possibility of any “interaction effects.” It’s time to repair this omission, and the main purpose of this paper is to take a rudimentary first step in that direction. In particular, it uses a general economic model that combines the four mechanisms to make predictions about their interaction effects — with comparative statics looking at the cross-partial derivatives. (So, if you really want to complain about equilibrium assumptions, here is your opportunity…)

  Anyway, my point here is just that I didn’t want the main practical point of the paper to get completely lost amidst more philosophical discussions about category schemes.

  Thanks again,
  Rich Makadok

• 4. Peter Klein  |  24 August 2006 at 11:27 am

  “. . .  more philosophical discussions about category schemes.”

  Like what I just wrote. Oops. :-)

• 5. Lasse  |  25 August 2006 at 7:48 am

  Hi Rich,
  Thanks so much for replying to my post. Let me begin by appolgizing. If I wasn’t totally clear on this, my post was not intended as a critique of your main points; the call for study of interaction effects, or the analysis of such effects that you contribute. I certainly agree that this shouldn’t get lost among discussions of category schemes, and certainly you should use the category scheme that best serves this main purpose. Nevertheless, your four theories inspired me to think about whether it would be possible to create a typology of theories of profit that that could be universally agreeed upon. I think this might be difficult, at least using criterea such as necessity, sufficiency, and independence. I tried to illustrate this using collusion.
  Regarding collusion and entry barriers, we seem to agree that while entry barriers are always necessary for profit under competition, they are some times sufficient (Cournot) and sometimes not (Bertrand, undifferentiated, symmetric, profit maximizing). Note here that I did not suggest that entry barriers are always sufficient, only that entry barriers can be sufficient (as in for example in a classic Cournot situation).
  BTW I love your aircraft analogy!
  Best regards
  Lasse

• 6. Richard Makadok  |  25 August 2006 at 12:03 pm

  Peter writes, “Disequilibrium profits, of the kind emphasized by Knight, would then be in a fifth category. (Such profits can’t exist under perfect competition, but they don’t require entry barriers, unless we classify asymmetric information, or differences in “judgment,” to use Knight’s word, as an entry barrier.)”

  I must admit that I am not terribly familiar with Knight, or with disequilibrium logic in general, so my comments here may be off the mark. If so, my apologies in advance.

  Please correct me if I’m wrong here, but my understanding is that the basic Knightian disequilibrium story goes something like this: Somebody recognizes a potentially profitable opportunity that everyone else is missing, and takes advantage of that opportunity before anyone else can do the same. If his move prevents others from following, then the profit may be sustained. If not, then the profit is fleeting. Folks might call this “disequilibrium,” but I would tend to think of it as simply exploiting a market inefficiency, not terribly different from Barney’s (1986 Man. Sci.) strategic-factor-market logic. It’s “entrepreneurship” in the original meaning of that word, which has absolutely nothing to do with creating new organizations. The original meaning of “entrepreneur” simpl referred to someone who inserts himself between two other parties in a transaction (like a producer and a consumer).

  As an example, I am reminded of the story of how Microsoft originally put itself in a position to dominate PC software. It did so by simply inserting itself between two other parties in a transaction — Seattle Computer Products, an unknown developer of a clone of the CP/M operating system called QDOS (for “quick and dirty operating system”), and industry giant IBM. IBM approached Microsoft to license its operating system for its new PC, completely unaware that Microsoft — which, at the time, only produced programming languages — did not actually have any operating system to license. But before the deal could even discussed, IBM swore Microsoft to secrecy with confidentiality agreements. Microsoft agreed to license its nonexistent operating system to IBM (non-exclusively, of course), and then bought the rights to QDOS from Seattle Computer Products for a mere $50,000, saying that it was for “a client” whose identity could not be disclosed due to confidentiality agreements. With a few minor modifications and some re-packaging, QDOS became PC-DOS, and then later re-packaged again for PC clones as MS-DOS. The rest is history, and the multi-billion-dollar return on on that $50,000 investment must be the highest ROI in history.

  Anyway, to me, all of this sounds a lot like what I’m calling “preemption” as one flavor of the committment-timing mechanism. The main difference is that I’m agnostic about what determines which player leads and which player(s) follow. It could be, as you suggest, asymmetric information,or asymmetric judgment, or it could be relationships (as in the Microsoft QDOS story), or simply luck. I don’t really care. Regardless of the conditions that led up to it, the proximate cause of the profit is still the act of preempting the opportunity. All the asymmetric information, asymmetric judgment, relationships, or luck in the world is irrelevant if you don’t actually make use of that information-judgment-relationship-luck by taking the necessary action to preempt the opportunity. For example, if Microsoft had simply told the folks from IBM, “Hey, we’re not really the right people for you to do this deal with. You should be talking to Seattle Computer Products instead,” then Microsoft would likely have remained an obscure footnote in the early history of the PC industry, and hardly anyone would ever have heard of Microsoft today (other than programmers who had used their programming languages back in the early days).

  Another indication that disequilibrium logic should fit into the preemption category is that models of preemption are not really true “equilibrium” models anyway. To me, the game-theoretic meaning of the word “equilibrium” (e.g., Nash) is that every player in the game takes into account its EXPECTATIONS of all other players’ moves when choosing its own move. In this sense, the players’ expectations equilibrate each other — everyone is responding, in anticipation, to what everyone else will do. Cournot equilibrium clearly fits this definition, with each player taking its expectation of the other player’s choice of output into account when choosing its own output. Likewise, Bertrand equilibrium also fits this defiinition, but with expectations of prices substituted for expectations of quantities. However, the classic model of preemption — a Von Stackelberg oligopoly — does not really fit this definition. Yes, in a Von Stackelberg oligopoly, the leader must take its expectation of the follower’s subsequent output choice into account when choosing its own output, but the same cannot be said of the follower. Because the leader has already committed to an output level by the time the follower moves, the follower has no expectation of any further moves by the leader. Thus, the equilibration of expectations goes in only one direction: Leader’s choice is influenced by expectations of follower’s response, but not the reverse. Therefore, one could argue, it’s not really a true “equilibrium.”

  Anyway, all of this is a long-winded and roundabout way of saying that the kind of disequilibrium profits you have in mind seem to me to be a special case of what I’m calling preemption.

  But, again, I say this with the caveat that I’m not very knowledgeable about disequilibrium logic in general, or about Knight in particular. So, I am certainly open to being corrected if my comments here seem off the mark.

  I also say this with the recognition that the commitment-timing version of my model totally ignores and abstracts away from any concern about what determines which player leads and which player follows — and that question may be of very central concern to disequilibrium theorists like Knight.

  I hope this helps. At the very least, it gives me some extra material for my discussion section.

  By the way, the paper itself is in a state of flux right now. I’m in the process of overhauling the model a bit, as well as the framing. So, I’m not comfortable distributing the current version. But anyone who is interested can send me an e-mail (Rich_Makadok@bus.emory.edu), and I will send them a copy of the new version when it becomes available (hopefully sometime next month).

  Thanks again for shining a spotlight on this research — greatly appreciated.

  Best regards,
  Rich Makadok

• 7. Peter Klein  |  29 August 2006 at 10:53 am

  Somebody recognizes a potentially profitable opportunity that everyone else is missing, and takes advantage of that opportunity before anyone else can do the same. If his move prevents others from following, then the profit may be sustained. If not, then the profit is fleeting.

  This sounds closer to Kirzner’s concept of “alertness” than Knight’s concept of judgment, but is not an unreasonable description. However, what I think Knight has in mind is not a world that is “mostly” in or “pretty close” to general equilibrium, with a few, potentially fleeting, unexploited profit opportunities available for alert entrepreneurs to exploit, but rather a world that is constantly in flux, that is tending toward, but never even approximating, general equilibrium. Because no one can forecast future market conditions with certainty, all profits (excluding monopoly profits) are “fleeting,” in this sense. To use the $20 bill analogy, in the rational expectations, Lucas-Sargent view of the world, there are no $20 bills on the sidewalk because if there were, someone would have already picked them up. In a looser interpretation, like Rich’s above, there may be a few $20 bills lying around, but not too many, and some people (“entrepreneurs”) happen to be adept at spotting them before other people. In Knight’s vision, as I understand it, no one knows for sure where the $20 bills are. Entrepreneurs make guesses about where they might be, and invest resources in digging equipment, labor, etc., and are sometimes successful at finding them.

  Incidentally, Rich’s point that the original meaning of the term entrepreneur had nothing to do with new firm formation, is exactly correct. Indeed, as Nicolai and I have argued in a few papers, this is true for all of the classic contributions to the economic theory of entrepreneurship (not only Knight but also Richard Cantillon, J. B. Say, Schumpeter, Kirzner, and others). These writers conceived entrepreneurship as an element of all market activity, not just the creation of new firms.

  As to whether Rich’s commitment-timing category encompasses Knight’s idea, well, my hunch is that the answer is no, but I have to think more carefully about it. Maybe some readers can help!

• 8. Richard Makadok  |  26 March 2010 at 7:51 pm

  After getting bounced at two other journals (mostly my fault for making bad choices in framing and organizing it), a slimmed-down version of this paper has finally been published at Management Science. Here’s the reference…

  Makadok, R. 2010. The Interaction Effect of Rivalry Restraint and Competitive Advantage on Profit: Why the Whole Is Less Than the Sum of the Parts. Management Science, 56(2): 356–372.

  Thanks,
  Rich

• 9. Peter Klein  |  27 March 2010 at 12:51 am

  Congrats, Rich!

• 10. Lasse  |  28 March 2010 at 1:35 pm

  Congratulations Rich. Well done and thanks for posting. All is well that ends well : -)

Trackback this post