Will Economics Imitate Life?

I had the privilege of attending last week a seminar organized around Matt Ridley's forthcoming book on human progress across time. Clive Crook briefly summarized it, and Robin Hanson did an interesting post based on his comments at the seminar. Robin implores that if we are going to theorize about humanity's history of technological progress and economic growth (to the extent we can measure it), we need to be precise about things. We cannot ascribe a central causative role to, say, the invention of writing if it doesn't fit the timeline of facts. 

The facts Robin has assembled--mentioned in his post and developed in this paper--tell a story of three primary "growth modes" in human history. These correspond with the advent of hunter-gathering, agriculture, and the industrial revolution, with each new mode being characterizied by a much faster rate of the doubling of world product than before. This is an enormous topic (and one, Robin pointed out to me, that should naturally be considered by a blog devoted to economic growth) and won't be resolved in one post, one essay, or one book (as the voluminous literature devoted to economic growth demonstrates).

But, Robin's provocative puzzle and the discussion around Ridley's new book made me wonder about the relationship between economic thought and human nature. Much commentary of late has attacked the conventional economic assumption of rationality, with a corresponding celebration of behavioral economics and its emphasis on irrationality. Yet each of these basically assumes that human thought defaults to deductive and inductive reasoning. People have a generalized principle (self-interest, rational expectations) that they naturally apply to every situation. Or, people encounter facts A and B and our hunter-gatherer brains irreversibly induce C.

Do our brains really work that way? Actually, no. As Ridley remarked in an earlier book in reference to the human brain: "No general-purpose induction here." We are rational, to be sure, but abductively so. What does "abductive" mean? I don't claim to be an expert, but the term itself originated with American philosopher Charles Sanders Peirce in his analysis of Aristotle and has variously been described as "inference to the best explanation" or the formation of explanatory hypotheses. It was also adumbrated by William Whewell and his "philosophy of induction" and has been explained as the method of reasoning used by Sherlock Holmes.

To greatly simplify, abductive reasoning is the constant creation of hypotheses to explain a set of facts or observation, usually precipitated by a new or surprising fact that does not fit the conclusion that deduction or induction would draw. This isn't confined to scientific investigation--it's how our brains function. Neuroscience keeps confirming that perception is inference, that our brains are theory-generating machines and, not surprisingly, some of the best work in this area has been done in the field of artificial intelligence, where researchers must figure out how the human brain works. In terms of vision, for example, there is a hole in the middle of every person's sight line, but our brains fill it by inferring from the surrounding images. In terms of daily life, we constantly draw inferences that seem to explain new or surprising situations and freely revise those inferences in the face of new information. (I suppose from this you could proceed to some sort of neuro-ontological hierarchy but I have neither the capacity nor inclination to pursue that.) Even economics has been touched by the revival of interest in abductive reasoning.

So what does this have to do with economics, economic growth, and human progress? Economics is the study of human behavior, which naturally implicates human thought processes--the closer one can get to how we think, the closer you get to more relevant theory. But Robin's questions about growth prompt larger issues: the question of the causes of growth and progress is really one of the most important that economists and social scientists can consider.

When people talk about economic growth, innovation usually appears on everyone's list as a causal factor. So what is innovation and what causes it? In exploring the causes of human progress, the brilliant economist Nathan Rosenbergn advanced the idea that the freedom to experiment lay at the core: "freedom to conduct experiments is essential to any society that has a serious commitment to technological innovation or to improved productive efficiency." Experimentation with ideas, tools, and modes of organization matters because "there are many things that cannot be known in advance or deduced from some set of first principles."

If abductive reasoning means anything, it does not mean knowing something in advance (the eventual logic of induction) or from first principles (the essence of deduction). In fact, Rosenberg's observations are reminiscent of both abductive reasoning and Donald Campbell's evolutionary model of how new knowledge develops: blind variation and selective retention (BVSR). Every idea or innovation begins as a variation, and they are blind in the sense that a true innovation is one whose consequences cannot be known or foreseen (or induced) in advance. The way it propagates is that it is selected from among other variations and is somehow retained. Perhaps abductive reasoning was missing or implicit in Campbell's thinking.

So if we're thinking about reasons for why human progress became so rapid in the last two hundred years, or relatively rapid around 10,000 years ago, is there any weight to be given to abductive reasoning? On its face, it sounds silly. As a matter of explanation, if the human brain works abductively, we cannot ascribe to it a unique role in growth. Is there a manner by which abductive reasoning could be externally institutionalized? I suppose the market might be such a mechanism because its development over several millennia worked to channel experimentation and provided selective retention.

I'm left at a seemingly dead end, it appears. The explanations for human progress and its increases must be endogenous to humans and human society (unless we're prepared to admit truly exogenous causes like alien visits into our analysis). The timing must be right, so an innate feature of the human brain likely cannot serve as an exclusive cause in 1800. Experimentation is attractive, but surely the human capacity for wonder is timeless? Institutions seem to be coming to serve as a default explanation for growth differentials, but an intriguing recent paper raises some questions on this front.

Part of me is tempted to find solace in circular explanations like feedback cycles. Robin's growth grooves post struck me as leaving options open only for discrete, one-way explanations. I could be misreading it, but it was only later (reading Ridley's Red Queen, by chance) that circular explanation occurred to me as an answer to Robin. That is, can we seek answers in the interaction among many different factors and the associated positive feedback effects they would generate, causes becoming effects becoming causes in turn? I realize that's far from a new idea in economics, but I seem to have been led there inexorably by my own semi-logical reasoning.

Comments

[Michael F. Martin]

The title of this post gave me a chuckle. Economics has always been about life -- lots of it. Moreover, the field itself is a microcosm for society at large. According to my reading of the history of the field, early economists were often informed and inspired by models from the natural sciences, including biology.

More specifically, there is a reason why Hayek wrote *Sensory Order*. Independent of Hebbs, he prognosticated long-term potentiation because he saw the mechanism replicated in large-scale, long-term group behavior. In the same way that neurons that fire together wire together, individuals who cooperate with each other, adhere into social units.

As to the specific question about the rational hypothesis, the following no longer seem questionable to me: (1) individual prefernces change over time; (2)there is no grand theory that permits us to predict future preferences from the past and present (see Conway Free Will Theorem); and (3) finite information costs would introduce errors into our predictions even if it were possible (see Arrow). Given these, even if we assume for the purposes of modelling that individuals are optimizing the fit of activities to prefernces, it seems inapt to call those attempts to make fits "rational." We might as well drop the misleading terminology.

[Derek Ozkal]

If I'm understanding it correctly, I agree with Michael's interpretation of what it means to say individuals act rationally. In my view, assuming that everyone acts rationally is fine, but does not really help you analyze anything on a macro level because individuals have localized knowledge, different value structures, expectations, etc. Knowing that everyone acts rationally doesn't allow you to aggregate anything and determine a group outcome like economic growth. For me if you want to understand economic growth, you just have to understand that eveyone acting in their own self-interest with a multitude external forces (institutions, legal frameworks, resources, conventions, incentive structures, gov't, etc.) creates what Hayek calls "spontaneous orders." Understanding the neurological basis for why humans act the way they do is interesting and could certainly help understand how the aforementioned external forces could be used to loosely "guide" human/economic activity (very loosely since this is a complex world and hence the reason the orders that emerge can't be predicted). As an extenstion of the point, I view reseaching economic growth as a qualitative analysis of how these external forces affect individual decisions. If you want to quantify it, well that's an inherently lost cause.

[Robin Hanson]

Feedback explanations are fine, but you have to say which feedback cycles are behind which growth modes, and then explain why those cycles weren't possible until roughly the times when they appeared.

[Michael F. Martin]

Robin,

I think the improvements in technology that permitted for more widely dispersed and aysnchronous communication are the most important mechanistic explanations for the closing of certain feedback loops. But I'm 100% in agreement on the need for more precision.

Also, as an afterthought on this post, I remembered that the name for the field of ECOnomics itself shares its etymological root with the study of life at a grand scale -- ecology