A Little Knowledge

A little knowledge goes a long way.  That’s the saying, correct?  Well you’d never know it by looking at economics.  It’s hard to find knowledge anywhere.

Now I’m not being facetious about the gaps in economic theory.  Let’s all give the discipline its due and say that it has done a masterful job of getting as far as it has based on the limitations it bounds itself with.

It’s just that sometimes those limitations are glaring and can stop someone in their tracks if they’re not steeped in the dark arts themselves.

The results of those limitations are often a stunning avoidance of topics that are crucial to understanding a real economy.  Or at least they’re crucial to someone less determined to be so willingly limited.

I know, this is all vague.  Let me explain.

I have been re-reading an excellent book by Cesar Hidalgo.  It’s called “Why Information Grows”.  It was first published in 2015, and is well worth the time it takes to read through it.  Hidalgo is another of those apparently endless line of MIT types who wander all over topics in an unconstrained manner.  In his case he sets out a wonderfully clear account of what information is — as opposed to the usual rather cramped version built solely on Shannon’s communication theory.

I read the book back when it first came out.  My eye had been caught by it’s subtitle: “The Evolution of Order, from Atoms to Economies”.  How could I possibly pass that up?  Obviously I was not alone it being captivated by that innocent line because Paul Romer, someone far more august than me, also read it upon its publication.  I know this because he blogged about it in 2015.

Hidalgo is able to provide answers that are salient in their absence in economics.  Interestingly Romer, whose blog urges us to think like scientists and uses the Hidalgo book as a source of what this might look like, seems to miss the really valuable insights for economics I find most useful.

Of course, I need to modify all that.  I have no evidence from Romer’s blog that he did, in fact, miss those insights.  I am assuming that miss only by dint of not seeing any subsequent commentary suggesting that Hidalgo’s message hit home.  And my feeling that the insights are valuable depends entirely on my own quirky view of economics.

With that said let me explain further.

It has always been clear to me that the process of production, and beyond that the process of growth in an economy, depends on the way in which various groups of people deploy knowledge.  Think of the infamous fudge called “Total Factor Productivity” that fell out of Solow’s masterful exploration of growth.  How on earth can centuries of economic thinking end up with two-thirds of growth being due to an unexplained phenomenon that was then given a suitably vague name and shunted off for future thought?

In my view it’s darned obvious that most of that inexplicable factor is the knowledge accumulated through time.  Anyone who has struggled through the introduction of a new product or service can tell you that acquiring the knowledge necessary to build and distribute it is the hardest part.  Getting the labor and capital is a breeze by comparison.  And  no amount of incentive will make that knowledge magical appear.  The knowledge has to pre-exist the incentive which can then induce the kind of reaction economics confines itself to.  Incentive structures without knowledge are sterile.  The analytical framework of economics takes the existence of knowledge for granted and simply describes the allocation of the other resources needed to combine with knowledge to create production.

Which brings me to a related pet complaint of mine: neither capital nor labor are fundamental resources.  If we are, as Romer urges, to be scientific about our thinking we have to recognize that both traditional components of the normal production function are horribly muddled and imprecise.  The great capital controversy of the early post-war years was about this imprecision.  What wasn’t discussed, as far as I know, back then is that labor is also a muddled term.

Labor, in a more accurate and useful depiction, is a combination of energy, as in muscle power, and skill or know-how.  So we can decompose it into energy and knowledge input into production.  If we do this it becomes obvious we need to learn how to quantify them both adequately.  Which is what I was hoping to learn that Romer had begun to attempt to do.  After all it would be leading us towards a more scientific understanding of production and a more accurate or useful production function.

If he has set out on that course I am unaware of it.  Perhaps  others have.  If not, all I can say is that economics remains unscientific and lacking in precision.  So much so that total factor productivity still exists as a talked about thing.

Then there’s a second issue Hidalgo attack with vigor: the oddity of the individual being the sole building block of economic theory.

The accumulation of knowledge has a few constraints.  The key ones are our individual capacity to learn and retain knowledge; and the geographical bias created by differential spacing of clusters of pre-existing knowledge.  The implication of these are dramatic for economics.  The second implies that development will take place more rapidly and with greater ease in location endowed with existing useful knowledge and know-how.  The first implies that as the complexity of production increases we reach a boundary of opportunity broached only by combining knowledge and know-how of groups of people.  We call these latter groups firms, and since industrialization most production has been within such groups.

Hidalgo actually introduces Ronald Coase in his chapter on the value of combining knowledge and its impact on the economy.  This, as you know, is music to my ears, although Hidalgo defers to economic terminology and repeats the anodyne phase “transaction costs” rather than suggesting we specify knowledge as a resource needing inclusion in production and that it is not simply the cost of acquiring knowledge that creates a need for firms, but that the combinatorial challenge presents an insurmountable barrier to production by individuals.  In other words, in complex production the firm is the standard model and production in the mode put forward by standard economic theory is an exception.  This is not a cost-based explanation for the firm, but a more fundamental knowledge-accumulation based explanation.

Now in an information dense society it is possible for an individual to acquire knowledge more readily.  This is where the concept of networks emerges as crucial to growth.  A firm is a centrally coordinated network, and the early proponents of digital technologies argued that such technologies would so reduce the cost of knowledge acquisition that it would become increasingly possible to undertake complex production without the administrative overhead of a typical firm.  The boundaries between what takes place inside a firm and what takes place within the auspices of a market would then shift away from the firm.  Or, at least, we would see different organizational structures  begin to emerge to challenge the supremacy of the firm.

While it is true we have seen a considerable effort to deconstruct firms in the last few decades, I would argue that the impulse for such an effort has come more from the incessant focus on delivering shareholder value that from the shift in costs associated in knowledge acquisition.  The origin, of course, of the shareholder value focus resides in the libertarian ideology of Milton Friedman rather than in any understanding of the core economics of the firm.

In any case, Hidalgo does us a service by pointing out where an effort to introduce a more scientific enquiry into economics might go.  As the challenge of combining knowledge into more complex and novel products and services increases, and as the economic explores across the opportunity frontier, as new knowledge allows it to do, the role of knowledge and know-how needs greater inclusion in our modeling of growth and in our explanations of production.

Right now economics falls far short of where it ought to be precisely because it treats neither information nor knowledge as fundamental resources and, instead sticks with the old, and definitionally confusing, categories of capital and labor.  If we are going to get into definitional discussions and arguments they ought at least be over the things that actually drive growth.  That would be scientific.

Addendum:

Let me add that this general thrust of argument is one where I would agree with Deidre McCloskey.  Whilst I take issue with her overly zealous defense of markets which in my opinion is a misallocation of virtue from the political domain into the economic domain, I would completely agree that the predominant driver of growth is the accumulation of knowledge.  

Obviously creating the conditions necessary for such an accumulation depends on the liberty of people to think and apply their thought in practical ways unhindered by institutional constraints.  I prefer to see that liberty first and foremost as a political concept which then allows a subsequent and dependent economic liberty to produce what we see as growth.  Markets, in the sense that economists discuss them, are political constructs for the purpose of undertaking economic activity.  Economics, then, is secondary to politics in the order of explanation of growth.