Diane Coyle

/ / , , , , /

Thinking like an Economist?

Thinking like an economist.

What a horrible thought.  Can you imagine anything more restrictive and less imaginative?  It requires you to disengage from reality and enter a world constrained by absurd assumptions, odd definitions, and a lack of foundation.  The entire edifice of economics sits, in all its glory, suspended in mid-air and relying on what the philosopher Daniel Dennett describes as “sky hooks” to hold it up.  Beneath it is only vapor and foggy notions of the phenomena it purports to describe and explain.  Just ask an economist to describe a market in detail and you will understand how vague the enterprise is.  

Perhaps I am wrong.  It takes a lot of imagination to conceive of what passes as economics as being a study of actual economies.  You need a really vivid imagination, for instance, to think that businesses actually calculate or apply marginal productivity on payday.  Or that the dense ignorance of total factor productivity is an actual thing and not simply a definition of how little economists know about how economies grow.  Or that there is something called an equilibrium calling out and pulling us all into a perfect quiet calm stasis devoid of the roughness or hurry-burly that makes theorizing reality so stressful.  Or … well you get the picture.

How about this: did you hear the joke about the economist?

Efficiency.

Well, you had to be there.

The obsession with concepts like efficiency and equilibrium have driven economists into a backwater of irrelevance and, dare I say it, incompetence, that “thinking like an economist” is something many of us might, rightly, try to avoid.  

And yet.

The failure of economics in the past few years and especially back in the early 2000s needs no recapitulation.  Of course I am exaggerating.  But there is something peculiarly sad about a discipline that once was regarded as the queen of the social sciences with its imperial pomp and invasion of its neighboring disciplines being reduced to a technocratic application of bits and pieces of mathematics.  Yet that’s where it is.  Lurking around in various corners of the administrative state applying its statistical techniques to potential public policy decisions.  It has become nothing more than math applied to large data sets.  Devoid of a moral compass it claims to be “neutral” and thus a valuable and impartial arbiter of policy.

Or, at least that’s the story told by Diane Coyle in her recent book “Cogs and Monsters”.  In her telling the entire macroeconomic enterprise has gone fallow.  Having failed in their grand enterprise of reining in reality and carving into simple little models, economists have given up.  They have abandoned the big picture and gone underground.  Nowadays they focus on applying their skills to data in an effort to extract patterns that might, or might not, then inform action at a more modest lower level of activity.  

Oh how the grand debates and battles of the past seem so long ago.  

Applied math is where it is at.

Actually Coyle wants us to think of this technocratic turn as an advance.  Economics, she argues, is emerging from the shambles of 2008 in this new form.  It has become the indispensable toolkit of policy makers.  It is re-emergent and will endure in this technocratic form.  

The problem with this is, as Elizabeth Popp Berman explains in her new book “Thinking like an Economist” the substitution of efficiency for equality or other moral values in the process of policy formation is anything but a neutral activity.

Berman gives us a tour through the steady invasion of policy making by the economics profession.  She exposes the insidious way in which “thinking like an economist” became an essential aspect of policy formation mainly because it allowed politicians to avoid having to make morally based decisions.  Such decisions, as we all know, are necessarily difficult and contentious.  The convenience of a supposedly neutral score keeper is thus compelling.  The policy maker can always retreat behind the suitably cut and tormented data and argue that, however unfortunate the result may be, at least it conforms to the hard reality of “facts”.  Thes facts being the consequence of math being applied to said data. 

You can hear the policy makers voice now:  how sad it is that the policy in question cannot be more socially comforting, but facts are facts.  Reality intrudes in the form of analysis.  And that analysis was carried out by “experts” who, the public can rest assured, express no bias.  It’s just the way the numbers turn out when we think like an economist.  In this fashion economists are like superannuated accountants.  Totting bump numbers.  Running the math.  Checking the decimal points.  Expressing no opinion other than the logic is impeccable.

Berman’s history of the systematic rise of the technocracy and the influence of “thinking like an economist” within it is a depressing read.  Both sides of the political aisle have, at different times, encouraged the adoption of the sort of analysis that economists are proud of.  Government agencies have been encouraged to convert to “thinking like an economist” in order to protect them against attacks on their potential value judgements.  Decision making has been sanitized of morality.  It has been sterilized.  So, instead of creating policies with a more morally based objective — such as tackling inequality — we now are reduced, by the thought processes of economics, to a much more limited target.  Everything needs to be “efficient”.

Whatever that is.

And so we arrive back in the fog.

Since economics doesn’t engage with reality and, instead, wraps itself within a warm fuzzy blanket of assumptions designed to fend off the difficult calculations needed to wrestle with all that chaos out there, it has produced a set of techniques entirely inappropriate for real world policy decision making.  And its obsession with efficiency is a salient example of this problem.  In order to maintain the tractability of its calculus economics has uniquely defined efficiency.  It has closed up the system it analyzes.  It adopts absurd assumptions and then explains its absurdity as being irrelevant to the efficacy of the calculus. The process matters more than the result.  Or the input.  As long as the logic is impeccable any amount of embarrassing absurdity can be tolerated.  

This is now the world of technocracy.  All these experts poring over giant data sets.   It is hardly value-free to argue that efficiency dominates all other criteria.  It is the argument that an economist might make.  

And knowing, as you all do, how economists make their decisions, you all ought to be worried.

Berman certainly is. 

/ / , , , , , , /

Symbiotic Machines and Our Ignorance

Well, what are we to make of all this?

Here’s something I came across in Diane Coyle’s recent book, “Cogs and Monsters”:

“A large range of factual subjects, evidently part of science and duly explainable (in principle) by empirical methods operating under natural laws, treats different kinds of inordinately complex and historically contingent systems — the history of continents and landforms, or the patterns of life’s phylogeny, for example — as not deducible, or predictable at all, from natural laws tested and applied in laboratory experiments, but crucially dependent on the unique character of antecedent historical states in a narrative sequence fully subject to explanation after the fact, but unpredictable beforehand.”

That’s Steven Jay Gould writing in 2003.

He gets it.

Does the economy fit into Gould’s set of “inordinately complex and historically contingent systems”?  Of course it does.  Indeed it’s one of the better examples.  We can always look back and untangle parts of the story as to how or why some economic phenomenon emerged or disappeared.  There are a wide array of perfectly valid economic points of view to cast backwards and articulate such stories.  Our arsenal is full of them.  But that doesn’t mean that any of them have future validity or that a trick played once will work out the same way twice.  Those natural laws so beloved of economists are simply observations that help explain the tracks in the sand we have left behind as we navigate the tides of technological, cultural, and cognitive change.  They are nothing more than a guide into the future and the impenetrable unknown.  

Of course it’s not quite as simple as that.

The enormous stability created by the path dependence of history dictates that deviance is unlikely to be too radical in the immediate future.  Further off is another matter.  

Who, for instance, could have imagined back in the early days of industrialization that our lives would be so totally intermingled with that of the machines we rely upon to deliver out current standard of living?

You are all, naturally, aware of Lynn Margulis’s contribution to evolutionary biology.  She argued that a major step along the way towards life as we know its was the symbiosis between various primitive life forms that ended up producing cells with nuclei.  The primitive forms still exist, but within a mutually beneficial greater whole.  They are no longer separable in the way they once were.

Do we need a theory of economic symbiosis?  Do we need to develop a narrative that organizes our steady journey into a machine mediated existence?  Alternatively, do we need to start discussing how our machines are becoming more and more to resemble us?  Our mutual dependence is quite clear.  Only our hubris stops us from recognizing the inevitability of a total merging in the future.  Where will the line between human and machine be drawn?  

Economists like to talk about “general purpose technologies” (GPTs).  The concept is one of those generalities that gets tossed around frequently in discussions of the history of economic growth and the contributions to that growth of various inventions.  GPTs are the holy grail of advancement.  Their effects permeate socio-economic activity to such a degree that each elevation in growth seems impossible with them.  The steam engine, electricity, the computer and now artificial intelligence are often mentioned as iconic GPTs.  But, perhaps, from the perspective of a less anthropocentric and very long term viewpoint the original and longest lasting GPT is the human.  That is to say that we deployed ourselves, our skills, our know-how, and our capacity to do work for millennia to solve the problem of survival.  It is only recently that we added other GPTs to augment that set of attributes.  We found an ability to multiply the effect of our own basic capacities and break free of the Malthusian trap.  In order to do so we entered into a bargain with our machines.  We designed them to serve us.  We own them.  We decide their fates.  We use them.  They would not exist without us.  And yet each of those statements has a negative.  As our dependence increased we began to serve them.  They came to own our ability to prosper.  They came to decide our fates.  They use us.  And we could not exist, at least as we now are, without them.  Where, nowadays, does the pragmatic line exist between the two species?  Especially as we are now designing machines that can think.  What next?  Machines that feel?

So, what are we to make of all this?

Perhaps that we need to re-invent economics in the manner Gould suggests.  As we march towards a true machine/human symbiosis a future economy is not predictable from the so-called laws of economics.  Those laws look backwards not forwards.  Ineluctable complexity is just that.  It cannot be avoided.  Let’s not pretend.  There are no laws sufficient for the purpose of prediction.  None.  So all economic arrangements are simply accommodative to our current circumstances.  Contingency is all.  

In a way this fits perfectly with the Hayekian view that Coyle references in her defense of so-called market forces.  She gives us the usual criticism, referencing Hayek, of centralized organization of an economy.  Since knowledge is dispersed and cannot be collated for central computation, we are told, decentralization is the best and only way to go.  She leaves it there.  At least Hayek left open the door to the rather obvious criticism: if we cannot compute centralized solutions due the dispersion of knowledge and the vastness of the volume of information, how can we then argue that any solution is any better than any other?  We cannot.  We have lost control of the situation.  We cannot identify any solution as being better, being in equilibrium, more efficient, or meeting any other qualitative criterion at all.  We simply do not know.  Everything is a guess.  Maybe based upon some fancy theorizing, but a guess nonetheless.  So markets cannot be called any more or less efficient.  We have, by using the Hayekian logic, acknowledged our ignorance.  To go beyond that ignorance is to venture into ideology.  

But you know all this.

Of course you do.

I ventured into this ramble because of a sequence of articles in the Financial Times.  The existence or absence of hubris is essential for understanding how we can deal with current issues.  In one article, Geoffrey Owen, a former FT editor, says the following:

“Shareholder-based capitalism, with competition between profit-making firms as its driving force, is the means by which society’s resources are used most productively for everyone.” 

That is a highly contestable statement.  Given our lack of true insight into the workings of the economy, and given the lack of integration of corporate theory into economics, by and large,  we simply cannot, and ought not, make statements like that.  We don’t know.  Indeed there is a steady accumulation of evidence that suggest shareholder value, when carried to the extreme, as it has been in recent decades, is greatly harmful to swathes of society.  So to sweep past that evidence and to make assertions that drop straight from the naive and idyllic self-serving assumptions of economists as if they were truths is misleading, and at a deep level purely ideological.

Contrast that with an article today by Martin Wolf who joins the lists as an advocate of monetary tightening in the face of today’s burst of inflation.  His entire article can be resolved into a single perspective: economies are too complex for any idea to be thought of as having enduring value.  The fashion in monetary theory ebbs and flows to reflect the current context.  What was good once is now bad.  And vice versa.  Wolf even trots out the old observation that technocrats have a foolish tendency to prefer being precisely wrong rather than roughly right.  He argues for humility and prudence on the part of central banks because of how little we know about the economy.  Apparently the humble and prudent approach is to squeeze the labor force to stop wages spiraling out of control.  I suppose neither humility nor prudence applies to profit margins which appear to be as wide as ever despite so-called cost pressures.   Economists seem to forget that despite arguing that inflation is “always and everywhere a monetary phenomenon” Milton Friedman gave himself an enormous caveat.  According to Wolf, he also said that money affected the economy with “a lag that is both long and variable”.  In other words he said he had no clue, but prefers to think it’s all about money.  In any case, contra the fashion in contemporary technocracy, prices are established by bureaucratic administration.  They are set to cover costs and achieve a target profit.  It is, as Kalecki explained a long time ago, the volume sold at this administered price, and the target profit, that will trigger a price change.

And yet there is Diane Coyle defending the advances in economics since the Great Financial Crisis, or GFC as she calls it, and the rise of economists as technocrats capable of taking a tough stand.  High on their elevated intellectual perch they can, according to Coyle, make those tough choices because they are not influenced by the vagaries of democratic influence.  In true technocratic style they can calculate and decide in society’s best interests.  Such Platonic perfection would be wonderful, would it not, were it possible?  

Personally I would prefer that they stop pretending to know something that they do not.  Frankly they should all take a leaf out of the critics of “total factor productivity” (TFP).  It is a measure of our ignorance.  Economics just doesn’t understand the drivers of growth in any coherent manner.  Why would it?  They are inscrutable. So TFP is just the odds and ends of ignorance.

Or, perhaps, it’s a measure of our symbiosis with machines?

Just a thought.

/ / , /

Economics 10Whatever

I have been less vexed about economics recently precisely because I have been focused on other, and more urgent, topics. The politics of our age are wondrously absorbing, if not a little disturbing. In any case I do try to keep an eye on what economists are up to. In that endeavor I came across a short article by Diane Coyle on the website “Project Syndicate”. If you want to know what is occupying the establishment it’s a good place to visit periodically.

And Professor Coyle is decidedly establishment.

Her article is couched as a book review/discussion of what ails economics. In it she attempts to refute the notion that economics is lost in a desert of its own making, as I and many others would suggest, but rather, she argues, it is vibrant and rapidly modernizing. She tries to persuade us that critics, like me, just don’t get what’s going own at the frontier of the discipline. There are, apparently, a load of breakthrough ideas some of which might make it into the textbooks sometime in the future.

Great.

But not great enough.

Two paragraphs towards the end of her article illustrate the issues:

“Behavioral psychology, complexity theory, agent based modeling and the like, along with historical narratives, and emphasis on institutions, methods such as randomized control trials and now big data and AI, are by no means a coherent new mainstream. It takes time to change curricula, and institutional inertia makes new approaches too risky and difficult for young economists seeking academic jobs and promotions”

What do we make of this?

On its face Coyle seems to be telling us that lots of exciting things are going on. After all throughout the article she is hinting at the new work people such as Andrew Lo are doing. Lo is the author of one of the books she reviews and she clearly is an advocate of his work. But, she tells us, he has been striving away at his non-mainstream efforts since 1986. Presumably that is most, if not all, his academic life. He is one of the people, Coyle lets us know, trying to incorporate advances in the human sciences into economics. He has, presumably, failed so far. After all not much of the modern economics curriculum is built on the kind of psychology and evolutionary theory Lo specializes in. By not much I mean really little. As in if you squint really really hard you might see a vague trace of it.

This is, of course, not Lo’s fault, but for Coyle to swat aside the critics by reference to a body of work that still sits resolutely outside the mainstream doesn’t bolster her argument. It weakens it.

Yes behavioral psychology is being taken on board in economics. Its about time. It is decades late. But the textbooks don’t yet reflect the change. They are firmly stuck in the past. Nor, in Coyle’s own words, is the profession apparently keen on updating itself. In her own words: it is difficult, if not impossible, for new ideas to form the basis of activity for young economists bent on climbing the professional ladder. Inertia seems to get in the way.

Doesn’t it always?

Perhaps in a generation or two we might finally purge the extraordinarily simple and naive version of psychology that still dominates economics. It was included not as an effort to understand human activity and its motivations, but to make analysis tractable. It was a prop for other ideas that needed to be given a footing. It wasn’t a result of serious study beyond that of the mid to late 1800’s when a description of utility and so on was deemed useful during the so-called “marginal revolution”. It was an afterthought, in other words, to a need to make economics appear more scientific. As psychology is has always been absurd, but economists aren’t psychologists, so they adopted absurdity with brio and dash in the service of their broader agenda.

Coyle goes on …

“Practicing economists outside universities do not keep up with the research frontier — although even here, useful tools such as behavioral economics, complexity theory, market design, and network theory are making significant inroads. Still, the economics taught in university departments, practiced in financial firms, and applied by policymakers remains heavily reliant on old-fashioned reductionist rational-choice models.”

Exactly, Professor Coyle. Exactly.

But what do you mean?

The first sentence implies that those of us out here in the real world are behind the times. This is something I would contest strongly. My experience is that the ideas mentioned in that first sentence are more likely to be thought about and tested outside a university than inside. At least in terms of a broad definition of economics. The business strategy world is steeped in them. It has been for a while now. I realize that economists still see business theory as a stepchild of economics not worth spending time on, but there’s some real action out there. And that action is having a direct impact on the economy. After all it is practical.

Then the second sentence appears as a non-sequitor. Having, fairly directly, disparaged non-university economics as being distant from the cutting edge, Professor Coyle then goes on to suggest that university departments too remain heavily reliant on old fashioned models. Worse still, policymakers are way behind too.

So where is the cutting edge?

Somewhere other than most universities. And certainly not in places like the Federal reserve Board, the World Bank, of the International Monetary Fund all of which are staffed to the gunnels with economists. Out of date very one of them.

Now, I don’t want to be unfair: Coyle has a point. There is lots of good stuff going on. It’s just that, as Lo’s career illustrates, it takes ages, if not generations, for it to appear in the mainstream.

So: if you are a student wanting to be on the cutting edge and you want to understand the economy — the real one and not the oddity that economists study — get out and about. Go study somewhere where the new ideas are being absorbed more rapidly. Or, alternatively, go locate other disciplines where those ideas are already old hat. You can then study the economy through the lens of contemporary thought without having to wait for the stodgy world of economics to catch up.

The odds are that it won’t.

/ / , , , /

Blind Leading the Blind

A few days ago David Ruccio posted an article titled “Crash and Learn” on the state of economics education. I want to elaborate a little further, although my usual skepticism on this subject does bridle a tad at the concept of economics education. Is that the same as “military intelligence”?

Anyway, in that article is this quote:

In Manchester, Diane Coyle also defends the basic methodology of economics. She says there is confusion among critics between microeconomics, the study of the behaviour of individuals and firms, and macroeconomics, the study of whole economies. Macroeconomics, she admits, “is broken”. But microeconomics is both robust and often verifiable with real-world data. What, she asks, can heterodox economists contribute to typical concerns of microeconomics, such as discovering the right mix of policy incentives to discourage obesity?

Therein, as someone once said, lies the tale.

Macroeconomics is “broken”. I quite agree. It’s nice to see an august member of the trade admitting that snake oil is snake oil no matter how clever the mathematics looks. Not that I blame the math. You can’t make something silly into something smart simply by expressing it in the formal language of math. If the root is rotten so is the formal outcome.

But that part I like more is that she goes on to laud the robust nature of microeconomics.

Wrong.

Micro, if anything, is worse than macro. It is so utterly disconnected from reality that it is incapable of anything other than talking about itself. Which it does loudly and proudly.

Let’s begin with a broad observation: one of the problems with the mess that is macro is that for decades we have been told to build on micro foundations. Not only does this deny the existence of any phenomenon that might emerge from higher level interactions between lower level phenomena, it assumes that micro has anything useful to say. Yet it is at the level of micro that much of the absurdity of modern economics has festered. Rational behavior in all its ridiculous glory is a tenet of micro. So is all the nonsense about perfection of information. To pick up a microeconomics text is to enter an alternative universe where real people have been replaced by packets of data especially created by economists so that their incentive models, constrained optimization methods, and other paraphernalia all work smoothly.

These packets of data, we call them ‘agents’, have no choice even in the artificial domain of rational choice. Why? Because the rationality impose upon them is uniform. It conforms to special rules. Those rules dictate the response an agent gives in any situation. Those responses are thus entirely determined, not by the agent, but by the combination of rules and circumstances. So much for freedom of choice. No one, in this setting, makes a whimsical choice. Indeed they never change their mind. Ever. Instead they plod on under the tyranny of a rule-laden simplicity obeying the orders of their economist experimental overlords.

This is not human behavior, the essence of which is the variety of response and the diversity of interpretation of identical circumstances. It is exactly that diversity that promotes robust and sustainable growth. It provides alternatives. It opens spaces for argument and thus learning. Most importantly, it is the groundwork for survival in the face of the endemic uncertainty that permeates life. The basis of all problem solving is the possibility of alternatives.

Microeconomics in its mainstream versions falls far far short of talking about this kind of real world problem solving; the kind that takes place in unconstrained open systems; the kind that is influenced by higher level pressures like social power structures bearing down on the agent; the kind that real people have to make when faced with limited or asymmetrical information; and certainly not the kind that deals with complex and manifold solutions.

No, microeconomics is deeply deficient and is thus the root error in economics. It is a badly misshapen replica of a small aspect of the totality of human behavior. But it is what economists play with. Dabbling, as some do, with institutional or incentive structures can only partly mask the error. The same goes for tricks like the introduction of transaction costs. It is relatively easy to use the techniques of economics as it is now taught to advise on the construction of contracts, incentives and so on. All you have to do is to limit the responses of your agents, give them a limited palette of possibilities, and to presume that they all behave the way they do in the utopia of the textbooks. All you need to do, in other words, is shackle them far from reality.

One more thing: the typical concerns of microeconomics have wandered so far afield from the core of economic behavior and into distant fields such as obesity precisely because of the failure to create a microeconomics that could deal with, explain, or predict that care behavior. All micro has become is a sophisticated technique. It is a technology or a method. It has no explanatory power in real world economies so it has been exported to deal with problems elsewhere.

Perhaps it has value out there. I do not know. But is is certainly not the robust portion of economics.

As for the aspect of microeconomics that purports to be the study of business firms. The less said the better. The entire purpose of business is to create and then exploit asymmetries of information. This is accomplished by the conversion of learning into routine and replicable activity so that cost can be driven out and profit enhanced. It has nothing to do with maximization or any of the other oddities that prowl about in the economics textbooks. But it does have a lot to do with the existence and history of social structures and networks, power relationships, culture, technology, geography, and the institutions that govern our lives.

No. Micro is just as big a bust as macro.