apophenia, or “why i’m ‘small government’, but not (just) an unbelievable asshole”

I learned a fantastic word a few weeks ago.  “Apophenia” is commonly understood (according to Wikipedia), as “the human tendency to seek patterns in random information in general.”

When I read the definition of apophenia, I (of course) immediately pictured Russell Crowe in “A Beautiful Mind”, with an office full of newspaper clippings, blackboards full of weird scribbles, and miles of red yarn strung up connecting them all together, stumbling around in his boxers looking all inspired and demented.  Upon further reflection, I recognized this same image as the one I had previously assigned to my schema for “human engaged in political thought.”

Here’s my thesis in bulleted list form:

  • There’s a lot we (generally decent people) want to do.
  • We tend to greatly kids ourselves about the scope of what we actually know how to do.
  • Trying out different ways of doing things has real costs.
  • We shouldn’t go incurring real costs unless we have some degree of confidence that we’ll get the results we intend.

In more detail….

There’s a lot we (generally decent people) want to do
You like social justice? (I like social justice!)  You like people not getting murdered?  (Ditto!)  Wish everyone could just get a fair goddamned shake in this beautiful/ugly/happy/sad world?  (Preaching to the choir!)  You wish everyone could live to 123 years of age?  (Well, I’d sign on for “could”, but I’m not sure everyone should have to.)  You wish 30,000 people didn’t die every year in the U.S. from gunshots?  (Me too — I assume we’re both wishing the number were lower, not higher.)  You wish no-one was ever hungry enough to be physically ill from it? (Count me in!)  You you had telekinetic powers (duh!)  Wish human existence didn’t have to include abject misery?  (Of course you do, and so do I.)

We tend to kid ourselves about the scope of what we actually know how to do
Apophenia!  Scientism!  Wishful thinking!  …give it a name!  Whatever you call it, human beings have an indefatigable ability to believe that we know those things we really want to know (regardless of the reliability/verifiability of that knowledge.)

I’m not sure how to say this without jumping down an epistemological rabbit hole and sounding like a wacky logical positivist, but the kind of knowledge I’m talking about here is that which can be used to make accurate predictions.  (The more accurate the predictions, the more powerful the knowledge.)  We have incredibly strong “knowledge” about some things (orbital mechanics, nuclear physics, etc..), and we have relatively weak “knowledge” about other subjects (pretty much anything that involves the collective behavior of groups of more than ten people.)

Now, I also happen to “know” that my children are the two most perfect beings in the entire universe, but that knowledge isn’t going to stop world hunger or put any satellites in orbit.  It’s an altogether different type of knowledge, expressing an altogether different type of truth.  That’s not to say that the this type of knowledge isn’t important in its own way, but it’s not important in regards to my ability to predict the behavior of the physical world.

Anyway, back to that relatively “weak” knowledge about the behavior of groups of more than ten people.  Our knowledge-building to date has left us at the point where we just don’t know much of anything about how large groups of people will behave (let alone how to make them behave the way we want.)  That’s not an indictment of science or the scientific method, it’s just an acknowledgement of the fact that this is a really difficult subject.

The current state of affairs is that trying to use public policy to control large-scale social phenomena is roughly equivalent to trying to control the details of next-year’s hurricane season by strategically herding butterflies for the six months leading up to it.  (We know that each flap of a butterfly’s wings has a cascading effect that will influence the path of a hurricane, but that’s about all we know.)

The NYC soft-drink ban is a good example of this.  Some researchers did some interesting research on individuals’ behavior when confronted with several different options for portion sizes.  These results were inappropriately extrapolated to serve as the basis for a proposed policy banning certain types of beverages over a certain size, with the stated intent of reducing obesity.  The problem, of course, was in inappropriately extrapolating research about what individuals in a laboratory setting did from what a population of millions in an almost completely un-controlled environment would do.  You can find wonks arguing both sides of what the outcome would be  (anything from NYC’ers finally fitting into their skinny jeans to soda-pop speakeasies and tommy-gun-wielding Coca-Cola distributors), but the most accurate assessment is a good old fashioned:  “we don’t know what would happen if we banned soft-drinks bigger than 16 ounces.”

Unfortunately, “I don’t know” are the three words that the human brain seems most predisposed to avoid at any cost.  The problem comes when we start replacing “this is our best guess  as to the result, and we don’t even have a way to estimate our confidence in our prediction” with  “this is gonna work!”

You probably snicker at the crazy shit that people believed in the middle ages, right?  Trepanation as a cure for epilepsy?  Mercury as a cure for anything?  (Other than a lack of mercury poisoning, that is.)  How quaint!  How hilarious!  How positively primitive of them!

Imagine how silly we’ll look in a few hundred years with our utterly demented notions that we can make people smarter with high-stakes testing, or that we can decrease gun violence by outlawing arbitrary groupings of cosmetic features on firearms, etc…

Imagine how hard they’ll be laughing when they read us describing something like the NYC soft drink regulation as an “experiment”, whose results are measured in obesity rates.

Trying out new ways of doing things has real costs

This applies to obvious “let’s go build some widgets” endeavors, but also to less obvious regulatory activity.  Want to ban soft-drinks greater than sixteen ounces in size?  You’re going to have to spend money enforcing the ban.  Soft-drink manufacturer’s are going to have to disrupt their supply chains.  Consumers wanting 17 ounces of soda, will end up paying for two sets of packaging.  Etc…

Everything that the government does has to be paid for by the governed, and  the more government “does”, the more the governed have to pay.

We shouldn’t go incurring real costs unless we have some degree of confidence that we’ll get the results we intend.

Whether you’re running a government, or managing your household budget, you have finite resources with which to achieve multiple objectives.  You prioritize; you allocate; you strive for efficiency.  Once you’ve covered the existential necessities, you find yourself deciding how to spend that last $10.  Do you rent an extra movie that week?  Do you spend it on lottery tickets?  Do you put it in savings?

Here’s what you don’t do:  you don’t take that last $10 and use it to buy sack of magic beans because you heard somewhere that magic beans will add ten years to your life.  And that’s what the vast majority of public policy is, magic beans.  That is, we have about as much reason to believe that banning certain types of beverage of a certain size will decrease obesity as the medieval barbers had to believe that “bleeding” a patient would restore his/her health by starving the angry vampire that was living in their stomach.

In review…

Anyway, that’s why I’m generally in favor of “small (minimal) government”, but not just because I’m a gigantic asshole.  It’s not because I’m a selfish prick who wants to keep my 99% all to myself.  It’s not because I think that government has no business caring about anything other than protecting my interests.  And it’s not because I think government is inherently  defective in the first place  (though I have my suspicions.)

The reason I’m “small government” is because I don’t think my government’s efforts at insert-policy-objective-here are qualitatively better informed than a bunch of witch-doctors running around decapitating chickens in hopes of appeasing the hurricane god and assuring a tranquil storm season, and I’m not OK with idea that everyone should be forced to ante-up when the witch-doctors want someone to pay for three million chickens.

And Some Good News:  We’re going to know a lot more in just an other generation than we do now.  Big data, bleeding-edge regression-analysis-type-maths, behavioral/hedonic economics, and exponential growth in computational power are already on the edge of giving us a whole new paradigm for knowledge-building in the social sciences.  In another few decades, we’ll actually be able to perform regression analyses that succeed in filtering out the insane complexities of large-scale social constructs and we’ll probably be a lot closer to having knowledge with the sort of predictive power I discussed above.

Of course, having that kind of knowledge/power raises a whole other can of worms related to the if/how/when it should be used, but that’s definitely another question.

(Also, my kids got into some toilet-related humor tonight, so I should have a more interesting post on that soon.)

UPDATE:  Adam Singer posted about this entry on his blog, Grey Politics.  He graciously described this post as well written and easy to digest, so his judgment is probably not to be trusted.  Adam also replied in the comments section of this post below.


6 Responses to “apophenia, or “why i’m ‘small government’, but not (just) an unbelievable asshole””

  1. Adam Singer Says:

    Interesting and valid points. I don’t quite draw the line to the conclusion you reach, though. There are not only “real costs” in putting new regulations or laws or guidelines in place but also real costs in not doing so.

    Suppose they would help and we delay enacting them in order to achieve a high degree of confidence they’ll work. We have spent the resources and time devoted to gaining that confidence and have lost out on the benefit we could have been getting during that time.

    There is a risk with acting too soon, of course, so a balance must be struck. And if an idea is clearly unlikely to effect real value we shouldn’t spend the time and resources to vet it.

    In a large part, it comes down to the nature of government. I believe Locke was more correct than Hobbes or Rousseau. The former gives government too much sway and freedom to oppress unchecked. The latter gives it too little to protect the citizens from each other (let alone external forces).

    The soda law was a very silly example. No – I do not think it should have been enacted and am glad it was struck down. It still is part of a larger and important conversation about how food and beverage manufacturers seek higher and higher profit margins at the expense of what it truly good for the general population. Sure, I can try to be informed and eat healthy but if the government didn’t have the power to force the companies to clearly label the nutritional contents my efforts would be blinded and fruitless, for example.

    • hlm Says:

      Thanks for the reply!

      The “cost of doing nothing” is one of the arguments I hear most frequently in favor of expanded government scope. I’ll happily stipulate that out of the 10 million things we could try to reduce the racial wage gap in the US, at least ten of them might actually have the desired effect. The problem is that 10 million is a conservative estimate on the number of things we COULD try, and throwing resources at a one-in-a-million chance that we’ll get what we want is no way to run a railroad. (Certainly not when those resources could be spent on a two-out-three chance of achieving some much-easier objective.)

      I expect that the most common response to my last line would be “so what; you think we should just STOP trying to solve the hard problems?!” To which I would say, “not exactly.” I’d say that if instead of throwing X amount of resources at what’s basically a one-in-a-million shot, throw X resources at basic research until we know we have a solution with at least a one-in-ten chance of working. I think THIS where most people and I part ways. I’m basically saying, “yes — there are things we care about deeply that we have NO IDEA how to fix (yet), and we need to learn how to admit that.”

      I suspect where you and I probably differ most is in our estimates of the likelihood of various policies in achieving their various objectives. That’s probably worth a conversation (or three) of its own, but for now, I’ll just say that I think that we (collectively) are victims of multiple cognitive biases which result in our DRASTICALLY over-estimating the predictive power of our knowledge in the social sciences (and other areas too, but SS has a huge overlap with government policy.)

      One minor quibble though…. I don’t think the soft-drink law was a very silly example; I think it was an excellent example of a very silly law 🙂 I like where you’re taking the discussion there though, though I’m too tired to follow up on that just now.

      • Adam Singer Says:

        Excellent point- I mispoke with “silly example”. As you state, I meant “silly law” (and, more completely, a perfect example of a law that illustrates silliness). But yes, my phrase was quite incomplete.

        As for the role of government, I don’t necessarily believe it must be constrained entirely based on what we fully and completely predict it can accomplish as the private sector and citizens are not so constrained. With the government thus tethered, it is unable to keep up and ineffective at responding to any new event until the moment has long past. We see this with the policies and regulations around digital information of all forms that are woefully out of date. What does “wiretapping” even mean anymore if you’re on a device using a wireless network? What if the network is unencrypted but you access a secure web site, like your bank?

        Yes, there needs to be analysis and understanding before action is taken but sometimes swift response is justified, even in the midst of unknowns. Of course, as the results of action become clear we need to then be able to adjust and revisit policies (this being a very difficult subject given how hard it is to even arrive at an initial action these days).

        • hlm Says:

          I agree that there is no need that government be constrained based on what we fully and completely predict it can accomplish (though I find the justification of “keeping up” with private actors somewhat specious.) I do think that government ought to be constrained based on what it has a (n as yet un-quantified) empirical basis for believing it can accomplish. I feel justified in wanting government to be held to a higher standard than private actors in this regard, because the government is granted coercive authority to enact its policies, and the resources the government uses to implement its policies have to be taken from private actors. That is, the guy with the gun telling everyone that they have to hop on one foot for two hours each day to prevent the next recession has a higher burden of proof than the guy who decides that he’ll jump on one foot for twelve hours each day for the same reason.

          My larger point is, of course, that in the vast majority of subjects the government wishes to intervene in, the results of action NEVER become clear, and that we engage in a collective apophenia when we convince ourselves otherwise.

  2. Billy Barnett Says:

    I think the soda law is a sufficient example. Behaviorism (which is what the soda ban is based on) demonstrates the government’s “business model” over the last hundred years to implement slow social change instead of guaranteed contractually agreed upon limited government. An organism that realizes unrestricted opportunity for slow growth has great opportunity to put the burden of that growth upon the population. One soda at a time.

    • hlm Says:

      Hi, Billy; nice to see you here 🙂

      I don’t think that the soda law was based on Behaviorism in any traditional sense; my reading leaves me under the impression that it’s based on the default bias and marginal analysis concepts which are rooted in economic disciplines (“behavioral economics” is a far cry from “Behaviorism” in psychology.)

      I was actually a bit confused by your next statement that behaviorism “demonstrates the government’s “business model” over the last hundred years to implement slow social change instead of guaranteed contractually agreed upon limited government.” I assume that you meant “structurally/constitutionally limited” rather than “contractually agreed upon”? Either way, the concept of a structural limitation of enumerated powers only applies at the federal level, and is (from what I’ve seen) extremely rare in subordinate layers (state, municipal, etc…), but let’s assume you’re referring to the federal level, in which case… it certainly does seem obvious to me that pretty-much every level of our government has indeed been involved in implementing various speeds of social change, but the interesting question there is “how/when IS that a legitimate function of government?” (Both in general, and within the framework of OUR federal constitutional republic.)

      Personally, I find it hard to argue the “government” does NOT have an interest in promulgating social change under some circumstances, but I have a much harder time articulating how to determine which social changes are apropriate/inappropriate objectives. (Balancing the rights of the minority against the interest of the majority and all that!)

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: