In April I wrote a long, winding review of Drift into Failure. I was very proud of it: in the space of several thousand words I visited complex systems, fuzzy logic and the usefulness of positivistic thinking.
It met with great silence; which is the lot of most blogs I suppose. But I had occasion one day to mention it to Helen Dale, who I host at the outstanding group blog Skepticlawyer.
Helen wrote back:
Now go read ‘Law, Legislation and Liberty’, where Hayek goes into exactly these issues in gory and grisly detail, right down to the argument that ‘constructivist raptionalism’ is derived from Cartesian thinking.
I found on further investigation that Law, Legislation and Liberty is actually a three-volume book, sometimes referred to as FA Hayek’s magnum opus. At the time I felt that, given the pressing weight of all the other books I want to read, Hayek’s opus was too far on the magnum side of things for me to justify picking up right away.
Then I found The Essence of Hayek, edited by Kurt R. Leube and Chiaki Nishiyama.
I, being lazy, decided that this was the ticket. At one stroke I could get a broad survey of Hayek’s thought and check to see if Helen was right about the connection between my review of Dekker’s work and Hayek’s writing (she was).
Still — it was slow going, for the reasons I originally outlined in this book review. I at least had the foresight to keep some notes.
So what is it about?
Actually, Hayek wrote and argued about a fairly wide range of issues. However not all of these have bearing on the original matter which led to Helen’s recommendation.
It seems to me the most important two themes running through most of the essays, lectures or book chapter were as follows:
- Spontaneous order (what I knew as “emergent phenomena”) is the only possible basis for successful societies and economies.
- Virtue and value are not the same thing; not only is the connection between reward and merit non-existent in a free market, any attempt to impose one renders the market system inoperable.
But that’s not all, of course. Various essays, book chapters and the like provoked enough thought and reflection that I’ve decided to break this book review into three parts.
Broadly speaking, I talk first of all about the theme of spontaneous order; then I’ll write in the second part about the problem of value and merit; then in the third part I’ll discuss some “miscellanea” (by which I mean I’ll lump together some disparate items which by themselves are, yaknow, only critical matters of political economy).
So, to the first theme — spontaneous order. For me, two entries in the book particularly stand out — Hayek’s speech about Dr Bernard Mandeville and one of his most famous papers, The Use of Knowledge in Society.
Dr Bernard Mandeville
Until this essay, I hadn’t heard of Mandeville, an 18th-century writer whose most famous work is The Fable of the Bees.
It was initially a poem, describing a beehive filled with avarice and greed. Mandeville takes great delight in skewering various professions including lawyers, physicians and government bureaucrats, painting each in a light of darkest venality.
Yet, in spite of its thoroughly greedy outlook, the hive is prosperous and powerful:
Thus every Part was full of Vice,
Yet the whole Mass a Paradise;
Flatter’d in Peace, and fear’d in Wars,
They were th’ Esteem of Foreigners,
And lavish of their Wealth and Lives,
The Balance of all other Hives.
Until, one day, the speech-making of a great moralist causes what might today most fashionably be called a paradigm shift:
But Jove with Indignation mov’d,
At last in Anger swore, He’d rid
The bawling Hive of Fraud; and did.
On the face of it, a wonderful change has overcome the hive. The lawyers no longer try to subvert the law, the physicians care more for their patients than their status, great swatches of the bureaucracy can be done away and the remainders are pious to a fault.
But then a strange thing occurs: the greed of bees, having disappeared, what we would today call the hive’s economy totally collapses. Factories close, businesses fold and mass starvation ensues.
The poem is ultimately meant to demonstrate that the “private vice” of greed supports “public virtues” — a wealthy, prosperous and powerful society. And that contrariwise, a society in which nobody was greedy — made entirely of the privately virtuous — would rapidly collapse into ruin.
The poem lay largely unnoticed until Mandeville wrote an elaborate commentary on it, which provoked a great deal of outrage: Mandeville was seen as directly undermining Christian ethics.
But the paradox remains. Hayek points to Mandeville as a critical figure in the history of ideas, because Mandeville is the first to clearly demonstrate how simple agents can form complex systems without central coordination.
My own approach to complex systems has been through studying them at university, first in terms of agent-based modelling, then classical systems thinking (feedback loops etc) and finally through an exposure to the nature-inspired computing / computational intelligence field.
Complex systems are characterised by “emergent phenomena”. They have particular behaviours which do not reside in any explicit listing of rules. Instead, the interaction of components gives rise to a behaviour which could not have been easily forecast only from a knowledge of the parts.
It was news to me that this concept emerged first to talk about social institutions and then, much later, was introduced into the natural sciences via Darwin. Typically the influence is thought to run the other way. For example, the standard agent-based or nature-inspired modelling approaches usually begin with relatively simple biological inspiration (a network of neurons, genes, an ant colony, a flock of birds) and then lever that into simulating all kinds of complex phenomena — including social structures.
The Use of Knowledge in Society
This is one of Hayek’s most famous essays. It’s famous because it goes directly to a central question of 20th-century political and economic thinking. Can an economy be centrally planned?
It’s important here to make sure that the question is clearly put. Hayek and others in the “calculation debate” were not discussing whether it would be difficult to plan. That is: they were not discussing the logistical boundaries of the problem. They were discussing whether it is possible at all, whether it is possible in principle.
It’s an important distinction. Suppose, for the sake of argument, that right now we had an absolutely 100% reliable way to calculate The Perfect Way To Produce Everything. Suppose furthermore (and this is an incredibly heroic assumption) that the computational difficulty of the universal equation scales linearly with the number of inputs.
Then the problem is one merely of having sufficient computing power. Assuming that, in this mathematic-socialist utopia that Moore’s Law still holds, then transistor density — and by proxy the amount of computations performable per second per mm of die space — doubles approximately every 18 months. No real-world economy grows so fast; so eventually computation will catch up with, and then overtake, our problem. Voila, the socialist utopia has arrived.
This is an example of where “is a planned economy possible?” is a matter of plausibility versus one of possibility. And in my argument above I made some generous assumptions. For example, I supposed that there exists some equation or algorithm that slows linearly with inputs. In computer science we call these O(n) algorithms and they are greatly sought-after. But computer scientists already know of hundreds of problems where no such algorithm is possible. And usually what is available is something ugly that, if you are foolish enough to demand the perfect answer, could happily consume all the matter and energy in the universe in its computation and still run out of time before the last proton has decayed.
Hayek comes at the problem from a different direction. He asks: where is the information for this One True Equation meant to come from? Supposing that we maintain, for now, the linearly-scaling algorithm we talk about above. We still need the inputs. But the set of all possible consumer products is very, very large. And then you need to consider all the intermediate products, products in various states of transformation, and so on and so forth.
Yet no such system, no matter how elaborate, can actually capture all of the information that each agent uses in their day to day work. It can embody neither knowledge of all current local conditions (eg that a welder’s blowtorch is not working at peak effectiveness) nor can it consider all the distributed know-how required for each person in the economy to do their job. A worker who welds every day has a mix of knowledge — some verbalisable, some kinesthetic — which is contingent on their long exposure to many different weld jobs. They know that for two steel bars the best weld is created by holding them just so and welding from this point on the join. Yet without this kind of local knowledge, the cental planner is not creating the perfect plan. There were, in theory, better alternatives that a local specialist could have identified. The central plan is, in actual fact, irrational — and perfectly rational central planning, in theory, impossible.
So much for central planning. But what about the alternative of free market operation? It might be argued that even an imperfect central plan may be more rational than the apparent randomness of the marketplace.
Here Hayek passes merely from demonstrating an impossibility to showing that the randomness of the market place actually allows coordination without requiring a central planner. Happily this means that the central planner, while being impossible, is also unnecessary.
How is this achieved? In the Hayekian model, the coordinating signal is prices. A factory manager, facing a change in the price of an input, does not need to know why or how it has come about. Was a mine closed? A factory has burned down? A new use discovered that drove up demand? It’s irrelevant to the factory manager: all she knows — all that she needs to know — is that the price went up and that she’ll need to cut down on that input to retain profitability.
Hayek is not the first to observe this per se. Adam Smith famously gave the analogy of an “invisible hand”, directing resources from one use to another without any agent being directly responsible for the decision. Hayek though connects the price mechanism with the way that knowledge about local conditions is widely distributed across an economy. The price system makes it possible for billions of people to coordinate their actions without knowing who is doing what.
Price is not the only signalling mechanism, of course. There are other ways for information to move through the economic system. Entire arts and sciences of management are built on adding more signals. But, by itself, the price mechanism is both sufficient and in a market economy necessary to settle all questions of what to produce, where to produce it, how to produce it, how much to produce, when to produce it, who should produce it … and so on and so forth. Many of these questions are formally unsolvable and yet market systems are able to produce sufficiently workable approximations that they outperform any imaginable alternative.
In a famously wasted opportunity, Wall Street villain Gordon Gekko said:
Greed, for lack of a better word, is good. Greed is right. Greed works. Greed clarifies, cuts through, and captures, the essence of the evolutionary spirit. Greed, in all of its forms; greed for life, for money, for love, knowledge, has marked the upward surge of mankind and greed, you mark my words, will not only save Teldar Paper, but that other malfunctioning corporation called the U.S.A.
This speech is silly. Gordon should have pointed out that greed is besides the point; someone is always going to behave greedily. What matters is not a quixotic quest to either expunge greed from, or elevate greed to, the moral pillars of humanity — what matters is how best to direct its energies to the common benefit and minimise harmful side effects. Hayek demonstrates that market systems — which connect greed to distributed knowledge through the price system — are pretty good at exactly that.