Need a name for a social science concept
October 4, 2011 2:59 PM Subscribe
Is there a "collusion number"? (A number past which a cartel is very likely to dissolve?)
This is a weird social science-y question - sorry if the explanation is awkward.
My friend and I were talking about cell phone carriers today just after the announcement from Sprint that they'd start carrying the iPhone. This idea started with a question: is four carriers enough to encourage real competition?
That got me thinking about oligopolies in general - situations where companies (or countries) stand in solidarity instead of breaking down and competing with each other a la-Prisoner's Dilemma.
With one or two members, it's pretty easy to stick together. But with each additional member, responsibility gets diluted, and it gets easier to cross the group.
Is there any concept in the social sciences of a collusion number?
This is a weird social science-y question - sorry if the explanation is awkward.
My friend and I were talking about cell phone carriers today just after the announcement from Sprint that they'd start carrying the iPhone. This idea started with a question: is four carriers enough to encourage real competition?
That got me thinking about oligopolies in general - situations where companies (or countries) stand in solidarity instead of breaking down and competing with each other a la-Prisoner's Dilemma.
With one or two members, it's pretty easy to stick together. But with each additional member, responsibility gets diluted, and it gets easier to cross the group.
Is there any concept in the social sciences of a collusion number?
One area to look would be antitrust law and economic theories of monopoly and market power. In fact, (on a related note) I believe parties opposed to the T-Mobile merger recently argued that T-Mobile was the market maverick (price discipline enforcer) in the US carrier market.
For a quick-and-dirty way of arriving at a possible answer to your four carriers hypo, see http://en.wikipedia.org/wiki/Herfindahl_index.
posted by Inspector.Gadget at 3:19 PM on October 4, 2011
For a quick-and-dirty way of arriving at a possible answer to your four carriers hypo, see http://en.wikipedia.org/wiki/Herfindahl_index.
posted by Inspector.Gadget at 3:19 PM on October 4, 2011
Hm. The Herfindahl Index measures how dominant a firm is in a particular industry but it's not clear to me if that's really what the OP is looking for.
posted by dfriedman at 3:30 PM on October 4, 2011
posted by dfriedman at 3:30 PM on October 4, 2011
It's a useful predictor of the likely competition in a defined market. From there, one can draw other inferences about likely cartel behavior.
posted by Inspector.Gadget at 3:32 PM on October 4, 2011
posted by Inspector.Gadget at 3:32 PM on October 4, 2011
There's not going to be a single number -- the answer will depend on many factors, like:
- the market pressure exerted by near-substitutes outside the cartel
- the elasticity of customer demand
- the degree to which the cartel is inflating prices
- the simplicity or difficulty of coordination among the cartel members
- the barriers to entry by new expected entrants
- the expectation of technological shocks to the industry in question
- intangibles like personal cameraderie (which might be reflected in the Dunbar number caek suggests)
- the degree to which it is possible in a given industry for cartel members to observe one another's prices and terms
A discussion of some of these factors (in the context of how to infer cartelization rather than how to sustain it) can be found in In re High Fructose Corn Syrup.
posted by foursentences at 4:41 PM on October 4, 2011
- the market pressure exerted by near-substitutes outside the cartel
- the elasticity of customer demand
- the degree to which the cartel is inflating prices
- the simplicity or difficulty of coordination among the cartel members
- the barriers to entry by new expected entrants
- the expectation of technological shocks to the industry in question
- intangibles like personal cameraderie (which might be reflected in the Dunbar number caek suggests)
- the degree to which it is possible in a given industry for cartel members to observe one another's prices and terms
A discussion of some of these factors (in the context of how to infer cartelization rather than how to sustain it) can be found in In re High Fructose Corn Syrup.
posted by foursentences at 4:41 PM on October 4, 2011
In antitrust, I've seen cartels with many more than four members/conspirators. They monitor each other to make sure they are acting anticompetitively and in furtherance of their conspiracy. I don't have a mathematical or other statistical answer, but I don't believe there is a magic number.
This is a different thing than the question of whether there are enough competitors in a market to encourage adequate competition, which is about market concentration and market power. I'm not entirely sure what you're asking -- can you clarify?
posted by J. Wilson at 5:18 PM on October 4, 2011
This is a different thing than the question of whether there are enough competitors in a market to encourage adequate competition, which is about market concentration and market power. I'm not entirely sure what you're asking -- can you clarify?
posted by J. Wilson at 5:18 PM on October 4, 2011
I think that, as with most social "science" concepts, while it is intuitively plausible that cartels grow more and more unmanageable as they get larger, one cannot talk about this with any kind of scientific rigor. There are simply too many factors in play to permit that. So I think the answer is "No," but only because there isn't (and can't be) a precise number, not because there's something wrong with the concept.
More generally speaking, there is no such thing as "The Market." Markets are diverse, and there is no good reason to think that the market in cars behaves all that much like the market in real estate, clothing, or crude oil in any but the most superficial ways. As such, market actors in various diverse markets may find that the characteristics of a particular market may permit or even encourage cooperation between larger numbers of market actors than the characteristics of other markets would allow.
Then there's the fact that once a cartel reaches a certain magnitude, it can become more resilient to maverick-type behavior. Take OPEC for example. Arguably the largest, most successful cartel in history. Yet not only has its market share decreased as reserves were found in non-member states, but member states frequently break ranks with the cartel to boost their own output. That notwithstanding, the cartel remains a significant force in the oil market simply because demand is quite inelastic (people aren't going to stop buying oil), there's very little that can be done to break it up (antitrust laws don't apply to sovereign states), and they control a large enough chunk of such a vast market that a little competition or maverick pricing isn't going to bring the thing crashing down.
posted by valkyryn at 5:51 AM on October 5, 2011
More generally speaking, there is no such thing as "The Market." Markets are diverse, and there is no good reason to think that the market in cars behaves all that much like the market in real estate, clothing, or crude oil in any but the most superficial ways. As such, market actors in various diverse markets may find that the characteristics of a particular market may permit or even encourage cooperation between larger numbers of market actors than the characteristics of other markets would allow.
Then there's the fact that once a cartel reaches a certain magnitude, it can become more resilient to maverick-type behavior. Take OPEC for example. Arguably the largest, most successful cartel in history. Yet not only has its market share decreased as reserves were found in non-member states, but member states frequently break ranks with the cartel to boost their own output. That notwithstanding, the cartel remains a significant force in the oil market simply because demand is quite inelastic (people aren't going to stop buying oil), there's very little that can be done to break it up (antitrust laws don't apply to sovereign states), and they control a large enough chunk of such a vast market that a little competition or maverick pricing isn't going to bring the thing crashing down.
posted by valkyryn at 5:51 AM on October 5, 2011
Response by poster: Thanks all! I think the discussion has helped to clear out my mental muddle a bit. To be sure, there isn't one "collusion number"--I'd think of it more as a situation-dependent "collusion threshold." I was wondering in particular (a) if there was a name for the concept, or (b) if some Ph.D. econ/social psych student had written an essay on identifying collusion thresholds and why they are where they are.
It also seems like the threshold might change if you were in a situation where there was one dominant player, which supports the idea of a threshold more than a number as well.
posted by Make Way for Ducklings! at 8:02 AM on October 5, 2011
It also seems like the threshold might change if you were in a situation where there was one dominant player, which supports the idea of a threshold more than a number as well.
posted by Make Way for Ducklings! at 8:02 AM on October 5, 2011
This thread is closed to new comments.
posted by caek at 3:18 PM on October 4, 2011