Monday, February 9, 2009

Metaphor and public policy: the danger of "primed pumps"

Warning, warning, danger, danger: severe geek alert.

First, let's take a short piece from one of my favorite philosophers that I love to read (often if only to disagree with his conclusions while admiring his writing), Daniel Dennett.

This is an excerpt from Consciousness Explained:

My explanation of consciousness is far from complete. One might even say that it was just a beginning, but it is a beginning, because it breaks the spell of the enchanted circle of ideas that made explaining consciousness seem impossible. I haven’t replaced a metaphorical theory, the Cartesian Theater, with a nonmetaphorical ("literal, scientific") theory. All I have done, really, is to replace one family of metaphors and images with another, trading in the Theater, the Witness, the Central Meaner, the Figment, for Software, Virtual Machines, Multiple Drafts, a Pandemonium of Homunculi. It’s just a war of metaphors, you say — but metaphors are not “just” metaphors; metaphors are the tools of thought. No one can think about consciousness without them, so it is important to equip yourself with the best set of tools available. Look what we have built with our tools. Could you have imagined it without them?

This is a critical point (about metaphors, not consciousness) that bears repeating: metaphors are tools of thought.

They set the patterns with which we conceptualize things, especially public policy issues.

Think about it this way: why do we have, in modern American political history, so many wars that are not wars?

War on Poverty. War on Drugs. War on Terror.

War is the most powerful public policy metaphor the government can use. It conjures specific mental images, emotional responses, and is particularly useful at mobilizing a sense of urgency, as well as "us versus them" thinking.

It is therefore also one of the most dangerous metaphors the government can use, because questioning the premises or conduct of a "war" is--at least implicitly--equated with a lack of patriotism or social solidarity. In other words, war as a metaphor not only mobilizes for the cause, but creates convenient enemies by demonizing policy opponents.

The Republican Party is the past master of this practice, but the Democrats have learned well at their feet.

All of this came to me as I was reading this post by Robert Reich. It will come as no surprise that I do not agree with most of Reich's economic views, but that's not the point of this post. The point is that Reich uses a very common metaphor in his article: priming the pump. [Actually, Reich uses the configuration pump-priming because of the grammatical requirements of the sentence.]

Priming the pump is a metaphor that pro-stimulus economists use to describe the process of injecting money into the economy to jump-start spending, which will then generate more jobs and production and .... so on ....

It is a common metaphor, and I want to make it clear that Reich's part in all this is simply that he used the phrase in a post I read just at the moment I was thinking about metaphors.

See, the pump-priming metaphor is one that is intended to get people thinking of the economy as a machine, with predictable responses to specific kinds of stimuli [like introducing fluid into a pump mechanism to create the proper mechanical conditions for operation]....

The metaphor implies several important things:

That the economy functions like a machine...

That those functions are well understood and can be manipulated or repaired with some certainty because cause-and-effect holds true...

That economists understand and model the economy effectively on a regular basis....

In point of fact, the economy is nowhere near being the simple mechanical system that the metaphor suggests.

The economy is a complex, non-linear system [it may even be a complex adaptive system]:

All good things come to an end and supremacy of the linear paradigm, characterised by utter certainty and predictability, was no exception. Einstein (1879-1955), Bohr (1885-1962), Schrödinger (1887-1961), Heisenberg (1901-1976) and Dirac (1902-1984) played a decisive role in pushing conventional wisdom within the natural sciences beyond the Newtonian limits that enveloped it centuries before. Later scholars working in different disciplines within the natural and life sciences demonstrated that uncertainty is an integral part of some phenomena; now described as being nonlinear.

In a reversal of the earlier trends at the beginning of the modern scientific era, discoveries in these sciences helped to initiate a process of change that is gradually introducing a measure of realism to expectations in the context of prediction and control of socio-economic events.

Nonlinear systems and processes do not present the familiar bell-shaped distribution associated with linear systems, where change is gradual and orderly and where measurements crowd together near an average value. On the contrary, Mandelbrot, and Gleick amongst others, discovered that in nonlinear systems change is more random and less predictable, and it involves discontinuities; rapid changes as opposed to smooth ones, and persistence; low for instance does not necessarily follow high.

Over time, a group of nonlinear entities attracted particular interest. These systems are variously described as being complex, because they have numerous internal elements, dynamic, because their global behaviour is governed by local interactions between the elements, and dissipative, because they have to exchange energy with other systems to maintain stable self-organised global patterns. In addition, when the stable patterns are capable of evolution the systems are also depicted as being adaptive.

For the purpose of this website discussion of social, political and economic phenomena will focus exclusively on their behaviour as Complex Adaptive Systems.

Linear systems are found at or near equilibrium. A ball bearing on the rim of a bowl is a classic example; it quickly settles at the bottom and that is that. Nonlinearity, by contrast, is exhibited by systems that are far from equilibrium

Complex Adaptive Systems are able to assume stable global patterns although they exist in conditions that are far from equilibrium. This is a key feature that merits a few words here. The second law of thermodynamics states that when an organised system is left alone it drifts steadily into increasing levels of disorder. A deserted building, for instance, eventually turns into a pile of rubble. After a while even the rubble disappears without a trace. Ultimately, a system cut off from the outside world will fall into a state of randomised equilibrium in which little or nothing of interest ever happens. Basically, disorder is a more probable sate than order.

Put another way, for a system to remain in an organised stable state, it has to exchange (dissipate) energy, or matter, with other systems all the while. That is the only option open to it to avoid succumbing to the destructive power of the second law of thermodynamics. Without the nourishing rays of energy from the Sun, for instance, Earth would drift into complete equilibrium, and therefore nothingness. Continuous supply of energy from the Sun keeps the planet in a highly active state far from equilibrium. The energy is absorbed, dissipated and used to drive numerous local interactions that in total produce the stable pattern that we perceive as life on Earth. It should be noted that local chaotic agitation is necessary to produce an overall stable pattern. In other words, organised complexity emerges from a mix of chaos and order.

A lot of jargon, but the point is more succinctly put here:

It is now possible to describe the traits that set Complex Adaptive Systems apart from other systems:

They have large numbers of internal elements that are lightly but not sparsely connected. The elements interact locally according to simple rules to provide the energy needed to maintain stable global patterns, as opposed to rigid order or chaos.

They have active internal elements that furnish sufficient local variety to enable the system to survive as it adapts to unforeseen circumstances. There are vast numbers of microstates inside the systems arising from numerous local interactions. There is, therefore, a high probability that at any time some of the microstates at least will find the prevailing conditions conducive to survival.

Variations in prevailing conditions result in many minor adaptations to the overall pattern of the system and a few large mutations, but it is not possible to predict the outcome in advance.

Predictability in Complex Adaptive Systems is limited to global patterns rather than the chaotic local details. Fundamentally, specific causes could not be linked to particular effects.

Pay particular attention to the emphasized sections of those last two sentences.

It is not possible to predict the outcome in advance.

Fundamentally, specific causes could not be linked to particular effects.

There is a growing body of rigorous academic literature that discusses the implications of applying the study of complex non-linear systems to social sciences, especially economics.

This study is increasingly indicating that economic systems as large as that of the United States cannot be manipulated through the implicit reliance on cause-and-effect policy as Keynesian economics requires.

Interestingly enough, most of the individuals studying this topic, when they are economists, are not Keynesians.

Or maybe I should say, are no longer Keynesians.

Point being: priming the pump suggests a much stronger assertion of the ability to control the economy by centrally generated inputs (command economy) than is warranted by the realities of our best understanding of how such systems function.

We need new metaphors for the economy, but we won't get them from people who believe slavishly in the ability of the State to achieve specific, predictable outcomes in a non-linear system.

Because to create them would be to admit that they really weren't that sure what they are doing....


Anonymous said...

Steve, an interesting piece, but you leave me then with only one solution to the current economic crisis, and that is: "Do nothing".

Isn't that the core of the Libertarian philosophy?

In the process of doing nothing, how many individuals will suffer grievously due to forces beyond their control? Is this that for which you wish?

In these complex non-linear systems, like the economy, like the weather, like the machinations of the universe, we are challenged to do the best we can to understand them anyway, so we construct models, which Keynes did for the economy, which Einstein did and now the string theorists are doing about understanding microscopic and macroscopic universal forces, which is what physicists, chemists, and the like are doing about understanding climate change.

So yes, we construct hypothetical models, test them, and once refined to a reasonable state project using the results, then repeat the cycle for further refinement of these systems which will never ever be perfectly refined, correct?

We are at such a point with a model of the economy, aren't we, rushed now because we are in crisis mode? There's no time for further refinement because people are suffering as catastrophic collapse seems a viable threat. This is not the first time, nor will it be the last as we muddle through this complex, non-linear global economy.

So we do the best we can with what we understand historically, applied to the best, most current model, which implies a choice to be made.

I suspect that it is the choice of the model upon which we disagree.

But people are suffering seriously, more every day, therefore we cannot "do nothing", as your piece seems to imply, can we? What is your model, Steve?

Perry Hood

Steve Newton said...

Good point Anon, about the piece appearing to leave nothing to do .... but nothing.

That, however, wasn't my intent. First I wanted to attack the way the problem has been framed, and then in a separate post (to keep the length of the thing beneath that of a Tolstoy novel) I wanted to examine what those researchers into such systems have suggested might work. Intended to write that today, but real life intervened.

Short verson (I will write the long version later): In CNLS (complex nonlinear systems) there are usually three aspects that can to exploited or manipulated with some prospect of success

1) Smaller local regions of the system will often respond to cause-and-effect tactics in a more predictable way; by analogy, this means that attempting a global (I mean "entire country" here, not "entire world") solutiion is likely to be less successful that tailoring solutions to smaller areas (either geographic or sector) of the economy. In a sense, trying a lot of little solutions rather than one big one. This is consistent with the rhetoric of Obama, Zandi, et al, but not with their acts.

2) CNLS researchers suggest that with economics building in "firewalls" between different parts of the system rather than creating a universal, single control or communications mechanism, will mitigate against local failure becomig systemic failure. If you look back for thirty to forty years you will see that we have (aside from the normal political regulatory arguments) tended to take down structural barriers between state banking operations, for example, in favor of more homogenous corporate entities. CNLS research would suggest that was a mistake, and some barriers to be re-erected.

3) CNLS research suggests that systems with fewer rules tend to operate more cleanly. I don't mean "fewer rules" here in the sense of "less regulation" or "less oversight." I mean "rules" in the sense of working patterned interactions (like the very simple rules of the computer game of "Life"). (I'm having a hard time explaining this right now because I'm home sick with a fever. sorry.) But the upshot would be that the merging of two many different investment/speculative functions into single corporate entities has created chaos rather than predicatability.

When banks are indistinguishable from mortgage companies are indistinguishable from brokerage houses (in other words when single corporations achieve complete vertical and horizontal integration), they become almost complete non-responive to regulation due to their size (they become CMLS in and of themselves). In a graphic sense, they become like a whole bunch of hurricanes appearing during storm season at the same time: the government has limited ability to predict what they will do, almost no ability to influence them overall, and is reduced to picking up the pieces.

The key here (and I suppose you will find this strange for a libertarian to be saying, but I try to be intellectually consistent in following the data) would be not to try to fix the current banking system as is, through government regulation or nationalization, but to take it apart into smaller pieces, with most regulatory activity fixed not on how the banks interact with their customers (which is where we generally think of regulation), but on keeping them from pursuing vertical and horizontal integration.

See? I told you it would get up to a Tolstoy novel.

Gregg said...

Here is another way of stating your point: The financial system has been allowed to become too tightly coupled and needs to be redesigned to make it more loosely coupled.

George Donnelly said...

"Isn't that the core of the Libertarian philosophy?"

No, the core is the non-initiation of force. Or some people say, self-ownership.

Here are some thoughts on what a libertarian might recommend be done:

Delaware Watch said...

"This study is increasingly indicating that economic systems as large as that of the United States cannot be manipulated through the implicit reliance on cause-and-effect policy as Keynesian economics requires."

Does your distrust of cause and effect analysis of economic systems mean that you now abjure your claims that printing money to deal w/ the deficit will cause inflation? It seems your non-linear analysis is a double-edged sword that could be used on the erstwhile Keynesian you cite as well.

Perhaps we need to be a little more complex ourselves about metaphors and, especially, what people mean when they use them, like according to Reich the benefit of the doubt that he doesn't hold that the economy can be described as a machine in all significant respects but only in some; unless, that is, you don't believe that there are any economic laws whatsoever.

Steve Newton said...

Good observation, but two points

1) Read my second post on repairing a CNLS

2) The economy is a CNLS; the government may or may not be (scholars disagree); but it is a much smaller system than the economy as a whole, and bankrupting it is far less difficult than scuttling the economic system

I know this is an incomplete answer and I will try to do better later, on the fly now but wanted you to know I had read and not ignored :)

George Donnelly said...

Doubling the money supply and spending a few bucks on pork are not analogous in their impact.

The former represents a doubling (or near-doubling) of a commodity people count on for the basics of everyday life.

The latter represents a drop in the bucket.aperali