This article about how bees and ants make collective decisions reminds me of my emergent stupidity theory:
So clearly it’s not enough to just put a bunch of dumb things together — how they are put together matters as well. But it at least offers the possibility that if you had a large enough bagful of Michael Moores (admittedly, it would require all of the burlap that the world will produce for the next century or so), you might have a chance of getting something intelligent as a result.
But to get back to my NASA example, I have a theory that the converse is true as well. You can aggregate a bunch of really smart things (like rocket engineers) and come up with something really, really dumb — an entity that would make decisions that no single individual among them would ever make, sans psychotropic drugs. Call it, if you like, the “committee effect.”
I’m not sure how to quantify it, but I suspect that it’s kind of like the rule for determining the resistance of a parallel network of resistors.
[Danger Will Robinson! MATH ahead!!]
If resistors are in series, that is, connected end to end in a long row of them, it’s easy to determine the total resistance — just add them up. So two resistors of ten ohms each become one resistor of twenty ohms when one end of one is connected to one end of the other, and the resistance is measured across the two free ends.
Parallel resistors, in which both ends of the resistors are connected to each other, so that the current flows through them all simultaneously, instead of first one and then the next and so on, has a different rule to compute the net resistance.
It’s: Total Resistance = 1/((1/R1)+(1/R2)+…+(1/Rn))
where the “R”s represent the individual resistances, and there are n resistors. In words, it’s the reciprocal of the sum of the reciprocals of the individual resistances.
For the example given above, it would be one over the sum of one-tenth plus one-tenth, or one over two-tenths, or one over one-fifth, or five ohms. So instead of doubling the resistance, as in the series case, we’ve halved it.
It can be shown (exercise left for the algebra student) that if all of the resistors are of equal value, the formula simplifies to the original resistance divided by the total number of resistors.
[End MATH]
Which is a frightening thought, if the same rule applies to my “emergent stupidity” theory. Assuming for simplicity that everyone in a government bureaucracy has the same I.Q. (it doesn’t change the answer that much if you allow variation, but assuming that they’re equal makes the calculation much simpler, as one can see from the formulas above), that means that the net I.Q. will be that I.Q. divided by the number of agency employees.
If you add the number of lobbyists and interest groups to the mix, you can drive it down another order of magnitude in value, to the point that it has the intelligence of a lobotomized fern (only slightly smarter than Joe Biden).
And my theory would seem to be borne out by the quality of decisions coming from, for example, the U.S. Agriculture Department, or the INS, or the State Department.
All of this, of course, is a long way of saying that I’m not encouraged by the prospects of merging several federal agencies and departments into a much larger (and probably dumber) one called the Department of Homeland Security, and then actually entrusting it with homeland security…
Just for those morons in comments who imagine that I was ever in favor of the DHS. I think the theory goes a long way toward explaining the hundred days, hundred f-ups that we’ve been seeing since the end of January as well.
I don’t buy it. It’s not a matter of numbers, but of organization and leadership. Some very large corporations run very well. Some small businesses — many, in fact, most — very poorly. (You don’t hear about them because they fail in short order.) What matters is the leadership and the incentives put in place for subordinates to follow. The problems in Washington stem from exactly that.
As for DHS. While bigger certainly doesn’t mean better, I think it provided at least an opportunity for two things: 1) improved coordination between the various groups under its umbrella, and 2) elimination of functions and groups duplicated among them. (I have no idea how well they’re faring at it.)
“a large enough bagful of Michael Moores ”
This gave me the image of some newly marketed version of Lorna Doones.
I recoiled at the image. I may not be able to sleep tonight.
I think it works well to consider every decision maker in a decision chain as a capacitor. They each add their own requirements, subverting the issue at hand.
The math works the same, though. The more layers of equivalently powerful decision makers you flow through, the worse the decision.
Great theory. While it may not be true, I still like the thought process. Almost as weird as I am (oh the horror!)
But seriously, I think leadership is the key. The problem is we’ve confused leadership for a popularity contest. We will pay for that confusion.
It’s really just a question of whether information is lost or improved each time it reaches a node, and (relatedly) whether the receiving node has the right incentives to act on that information in a way that benefits the purpose of the whole.
Free markets produce more efficiently than government because the price signal both communicates relevant information (“Demand exists here”) and provides incentive (Profits).
DHS isn’t necessarily bad. Organizational barriers to sharing information harm security. The problem is that the “tip of the spear” isn’t authorized to act intelligently on the information it receives. Greater independence of action (closer to what a free market agent enjoys) combined with superior incentives could overcome much of these difficulties, just as squad-level tactics improves military effectiveness.
There is no umbrella in this country bigger than Executive Branch of the Government of the United States of America. all that was accomplished by creating DHS was adding layers of middle management so that when the president steps up and says, “It happened on my watch, I take full responsibility,” most people will still not hold him responsible.
To summarize what has been said here, there are two drivers for “stupidity” in a large bureaucracy, namely diffusion of responsibility with corresponding lack of incentive for risk-taking, and complexity and inefficient communication.
I dislike the characterization in terms of IQ. That quantity doesn’t make sense from the point of view of a bureaucracy. Nor does the author explain why reciprocal averages should be the means for calculating some sort of intellectual capability of a bureaucracy.
Nor does the author explain why reciprocal averages should be the means for calculating some sort of intellectual capability of a bureaucracy.
Well, I was the author. And the column was at least somewhat tongue-in-cheek.
I think what you’re getting at, Rand, is what people who do this stuff call “universality classes.” Our most fruitful theoretical approaches to what you’re calling “emergent” behaviour suggest strongly coupled many-body systems, which is the only place these things happen, are best understood by grouping together systems with similar long-wavelength mechanisms of interaction, and glossing over the unimportant details of the interacting degrees of freedom at their base. These are the “universality classes”: when you fuzz over the local detail, systems in the same class behaviour similarly.
It seems quite likely that we share a universality class with ants and bees, in the sense that when we mostly follow instinct, we are significantly more intelligent than we are as individuals. This is hardly surprising. It’s not like we’re the first species in which Mother Nature has tried out very strong social interaction. She’s had 300 million years to perfect the DNA wiring.
On the other hand, we are the first species on which she’s tried out conscious reasoning, and the experiment is quite recently started. I think there’s a decent chance our conscious cooperation falls into a different universality class, one in which the emergent behaviour is, as you say, towards incompetence, stupidity, and disaster.
I mean, as a rule, the quality of outcome of conscious cooperation tends to diminish strongly with the number of people involved. Two people who know each other well can often make excellent joint decisions. Five — rarely. At best, with iron discipline, e.g. in the Special Forces, they make decisions no worse than an individual would. Above that, things fall apart rapidly. The quality over time of decisions made by groups of 50 or more tends to be appallingly worse than the quality of decisions made by even random individuals. This is where stock and tulip and real-estate bubbles come from, as well as Jonestown, Gulag Archipelago, and Buchenwald horrors.
Arguably our best instincts know this, and warn us against the folly of trying to consciously cooperate over too large a scale. This is perhaps why the worst excesses of collectivism are relatively rare, despite it’s powerful appeal to the conscious reasoning mind. Imagine how much more we could get done if we pool our efforts, cooperate, each consciously contribute our wisdom so that it adds up, like resistors in series, to an Einstein squared brilliance that can solve every problem the universe poses!
It takes a fair amount of experience to realize how very wrong that reasoning process is. That makes it surprising how little it takes hold among the young and inexperienced. I would tend to credit instinct. Mother Nature has probably armored us against the deadly mental virus by some inborn degree of narcissism and alienation. Our deep-seated mostly unconscious urge towards personal glory tends to make us uncooperative ants that throw away their programming when it comes time to build the One True Hive Mind that could rule the Universe.
Shorter Carl: collectivism can only work if one ignores or disbelieves in human nature, which is why collectivists do that.
Carl, when are you going to get a blog that we can link to for this stuff?
Does this also apply to Businesses?
When two firms merge do they add stupidity?
Would that explain why most mergers result in lost shareholder value?
When two firms merge do they add stupidity?
Of course it does! There is very little difference between big business and government, anyway. The same decoupling of action and effect happens.
Of course, most large business mergers are also the result of the emergent stupidity that we are discussing.
(BTW, only large – large business mergers fail on average. Large – small mergers are normally successful.)
Have can you not have linked to this? It’s the poster version of your entry and one of my favorites from there.