Wednesday, March 24, 2010

Free Minds, Free Markets, and Complex Adaptive Systems (Part 1)

(Howard Pyle, Soldiers)

The man of system, on the contrary, is apt to be very wise in his own conceit; and is often so enamored with the supposed beauty of his own ideal plan of government, that he cannot suffer the smallest deviation from any part of it. He goes on to establish it completely and in all its parts, without any regard either to the great interests, or to the strong prejudices which may oppose it. He seems to imagine that he can arrange the different members of a great society with as much ease as the hand arranges the different pieces upon a chess-board. He does not consider that the pieces upon the chess-board have no other principle of motion besides that which the hand impresses upon them; but that, in the great chess-board of human society, every single piece has a principle of motion of its own, altogether different from that which the legislature might choose to impress upon it. If those two principles coincide and act in the same direction, the game of human society will go on easily and harmoniously, and is very likely to be happy and successful. If they are opposite or different, the game will go on miserably, and the society must be at all times in the highest degree of disorder.
-Adam Smith

The Big Man vs. The Market

This blog entry and the one that follows are intended as explorations of the continued attractiveness of heavy government intervention and regulation of the economy, despite voluminous evidence that intervention damages the market pricing mechanism and creates distortions. I contend that many of us resist the evidence because two psychologically-appealing features of politics-based economic activity give command-and-control an intuitive attraction and thus the privilege of serving as our default state, while we permanently put markets on probation and abandon them at the first sign of trouble.

For a self-interested political operator, the advantage of a mixed economy is that the market portion can be blamed when things go wrong, while the politico can attempt to take credit for the prosperous periods (claiming to have invented the internet, etc). Indeed, the proposed ability to blame "animal spirits" is part of the allure of the Keynesian gigolo.

We begin with a quick discussion of the two stylized ways that a group can choose to allocate scarce resources by providing selection pressure for economic ideas that arise from two different sources: 1) voluntary exchange between individuals all seeking value (the Market approach), and 2) political processes that, one way or another, impose involuntary exchanges between at least some of the individuals and others (the Big Man approach).

In The Origin of Wealth, Eric Beinhocker gives us an overview of the situation:

Over its history, humankind has evolved two methods of economic selection: Big Men and markets...In the early days of the economy, the selection process was fairly straightforward---survival. If your Business Plan for combining Physical Technologies (e.g. bows and arrows) and Social Technologies (e.g., hunting party) under a Strategy (e.g., hunt impalas by the river) was successful, then your caloric revenues were greater than your caloric expenses. This caloric profit enabled you to do things like invest in children. Calorically profitable Business Plans then had a higher chance of being replicated by attracting more participants, and propagated over time by being adopted by the next generation.

As society and the economy grew more complicated, however, the feedback loop of selection became less direct, with intermediate, socially driven selection cropping up. The first collision between selection and society undoubtedly came when the first Big Man said, "Let's give this nice, fertile plot of land to my third wife's cousin (who is a lousy farmer) instead of to Mr. X (who is an excellent farmer)." We can safely assume that political meddling in economic affairs is as old as both politics and economics themselves. Such decisions favoring poor Business Plans (the cousin's) over good Business Plans (Mr. X's) do not last long in an environment where everyone was on the edge of survival---if the Big Man did this often enough, then either the Big Man's tribe would perish under his leadership or he'd be overthrown in a revolt. But once a society had crossed the survival threshold (particularly after the advent of settled agriculture), such social short-circuiting of the selection process became not just possible, but ever more likely as the group grew richer.

If a tribe is generally surviving and the Big Man's graft, corruption, or incompetence isn't life-threatening, then relatively few people may even be aware of the additional wealth their tribe is giving up.

(Big Man vanity projects made possible by settled agriculture and resulting high population densities)

The Fitness Function

Evolution is cleverer than you are.
-Orgel's Second Rule

I can now introduce the concept of the fitness function, which is critical for understanding everything from Darwinian evolution to modern computational methods such as genetic algorithms (a future blog post on trading system design will get more into GAs, simulated annealing, and their potential benefits and weaknesses in that particular niche application). For our purposes here today, the fitness function can be thought of as the scoring criteria that is used to determine who in the economy will live and who will die. Once the fitness function is in place, individuals and groups within the economy will search the "solution space" for strategies that maximize this function within whatever additional constraints they face.

(form follows function: Great Whites appear to be perfectly designed for what they do because their evolution has been shaped by a blind selection process to optimize a fitness function within a specific trophic niche)

In a market-based economy, the fitness function is the ability of a firm to sell enough of its product to cover its costs and provide a return to its shareholders. Financial analysts who build spreadsheet valuation models for companies know that the single most sensitive variable in the entire model is almost always the selected trajectory for sales growth. In a market, the way you maximize your fitness function is to sell things that a lot of people want to voluntarily buy from you.

In a politics-based economic system, however, the fitness function is the ability of a firm to win favor with Big Men. Big Men, who may or may not even use any of the companies' products, decide which companies will survive and which ones will not. The capacity of consumers to make voluntary exchange decisions is constrained because Big Men have taken (or been given) the power to set prices and wages, either directly or through subversive means. In a Big Man economy, the way you maximize your fitness function is to sell things that the Big Man wants you to.


One of the great accomplishments of Traditional Economics was to show in effect that the fitness function that markets attempt to satisfy is the overall welfare of the people participating in them. In a Big Man economy, a business lives or dies by political favor. In a market-based economy, a business lives or dies by whether its customers like and are willing to pay for its products and services. In a Big Man economy, resources are directed toward the ventures that best line the pockets of the Big Men. In a market economy, resources are directed to ventures that make the best economic use of them.

(a "fitness landscape" illustrates the possibilities for satisfying a given fitness function. The higher the elevation you achieve, the higher your fitness function score)

The crucial element in the Big Man economy is compulsion---economic decisions regarding resource allocation are forced upon subjects. In Knowledge and Decisions, Thomas Sowell examines the resource-allocation efficiencies of the two types of economic systems and adds the distortions created by the Big Man/political system's threat of force:

The element of force is crucial to the distortion. The knowledge transmitted by voluntarily chosen prices conveys the terms on which various forms of mutual cooperation are available. The knowledge transmitted under government price constraints reflects the desire to escape punishment, and the knowledge conveyed by such prices does not reflect the full array of options actually available to the economy.

The powerful insight that market prices are a knowledge-conveyance mechanism that reveals where pockets of cooperation can be found---goods and services clearing the market through voluntary transactions taking place at those price levels---is perhaps the most important in all of economics. It is a centerpiece of Hayek's work, and it is by meddling with the price mechanism that Big Men seek to shift resource allocations away from where a free market would go and towards where the political leader and his supporters would like them to go.

For example, taxing one group to pay for programs that benefit another group is simply a shuffling exercise. If the funds had been untaxed, they would have been deployed in the market via consumption or savings/investment, and someone on the other side of the transaction would have been made better off. Now that particular transaction will never take place because the money was appropriated for a use that a Big Man felt was more useful (to society? Possibly, but certainly more useful to himself, as he is a self-interested player as well). The scope of options for voluntary transactions is made smaller in favor of increasing the scope of involuntary transactions.

Sowell uses the example of populist minimum-wage laws to punctuate this problem:

Minimum wage laws likewise prevent transmission of knowledge of labor available at costs which would induce its employment. By misstating the cost of such labor, it causes some of the labor to be unemployed, even though perfectly willing to work for wages which others are perfectly willing to pay...the law itself does not guarantee that ANY wage will be paid, because employment remains a voluntary transaction. All that the law does is reduce the set of options available to both transactors...What is perhaps more surprising is the persistence and scope of the belief that people can be made better off by reducing their options. In the case of the so--called minimum wage law, the empirical evidence has been growing that it not only increases unemployment, but that it does so most among the most disadvantaged workers. This undermines some of the key assumptions of the price fixing approach.

Big Men vs. Markets in World Economic History

The curious task of economics is to demonstrate to men how little they really know about what they imagine they can design.
—FA Hayek

In his economic history of the world, A Farewell to Alms, Gregory Clark presents us with this graph:

The graph displays real income per person in England from 1260 to 2000. Note that income varied little---outside of mild statistical wobble caused when the aftermaths of mass deaths allowed for brief periods of rising incomes---for about six centuries, rising in 1800 and then rising very sharply in the second half of the 19th century with the broad effects of the Industrial Revolution.


But for the majority of the English as late as 1813 conditions were no better than for their naked ancestors of the African savannah...even according to the broadest measures of material life, average welfare, if anything, declined from the Stone Age to 1800. The poor of 1800, those who lived by their unskilled labor alone, would have been better off if transferred to a hunter-gatherer band.

An even more startling graph is revealed in the book's introduction, where Clark shows that, on a global scale, income per person changed very little since BCE 1000 (the graph can easily be extended to BCE 100,000 with virtually no impact on the results) . Once again, the "Great Divergence" erupts at the time of the Industrial Revolution; before this, the growth in real incomes is essentially flat.

Clark continues:

Thus the average person in 1800 was no better off than the average person of 100,000 BC...there was no upward trend (in) income per person---the food, clothing, heat, light, and housing available per head. The quality of life also failed to improve on any other observable dimension. Life expectancy was no higher in 1800 than for hunter-gatherers: thirty to thirty-five years. Stature, a measure both of the quality of diet and of children's exposure to disease, was higher in the Stone Age than in 1800. And while foragers satisfy their material wants with small amounts of work, the modest comforts of the English in 1800 were purchased through a life of unrelenting drudgery.

Why did this story unfold the way that it did? Most economists would cite changes in institutional arrangements and technological innovations that led to dramatic increases in what is termed total factor productivity. Clark's thesis, which is controversial and sometimes criticized as being an example of Social Darwinism, is that cultural changes in England in the 1800s created values that are prized in a capitalist, market-based economy and allowed humanity to escape the "Malthusian Trap" that had dominated for thousands of years.

Regardless, no one can argue that Big Man resource-allocation strategies, incentive packages, and political techniques were not given their fair chance. By Clark's reckoning, they had almost 100,000 years to work out their bugs before market-based economies and individual entrepreneurs harnessing new ideas and innovations to relatively free capital markets came about.

It isn't that Big Man systems fail outright so much as it is just that the fitness function being optimized really only benefits the Big Man and his friends. As Beinhocker puts it, "The immense mansions and palaces dotting the world, from grand French chateaus to the Hermitage in Russia, that delight tourists with their extravagant displays of riches are testaments to the effectiveness of economic evolution in maximizing the fitness function of Big Man wealth."

(Hermitage, St. Petersburg)

What we can see from human economic history, then, is that Big Man systems do create increasingly powerful forms of wealth as the population grows, but this wealth is nearly completely sapped by those with political power. Over millenia, we see that the personal net worth of political leaders is highly and positively correlated with the concentration of political power, while the population-at-large sees no rise in individual income unless market-based economies are available to incentivize decentralized technological experimentation and innovation.

Why We Still Believe in Big Men and Don't Trust Markets

With the evidence so completely stacked against the efficacy of the Big Man form of centralized economic planning by political elites, one cannot help but wonder why these systems, or at least mixed-systems that still feature some kind of heavy role for active government management of the economy, are still abundant. I believe that, at the psychological level, there are two primary reasons for this, and I will term them the Appeal to Intuition and the Appeal to Ambition.

I would describe the two appeals as follows:

-The Appeal to Intuition suggests that a system which we depend on to reliably produce complex, critical goods must either be designed from the ground up (this argument has largely been destroyed by experience, but it still persists in some circles), or at the very least have someone in charge of it.

-The Appeal to Ambition suggests that a system which offers strategic control and leverage points also offers opportunities to gain the power to harness the coercive powers of the state to one's own visions and ambitions.

My plan is to spend the rest of today's blog dealing with the Appeal to Intuition, and then attempt to contend with the Appeal to Ambition next week.

The Appeal to Intuition: Economies Must Be Intelligently Designed

Richard Dawkins, Daniel Dennett, Michael Shermer, and a number of other intellectuals have articulated the reasons why "Darwin's dangerous idea"---that great complexity and utility in the natural world arises from internal processes and does not require an external uber-designer, presumably of magical origin and status---is so actively resisted by unaided human intuition.

(Darwin's Bulldog: the great Richard Dawkins)


It is almost as if the human brain were specifically designed to misunderstand Darwinism, and to find it hard to believe...a (respect) in which our brains seem predisposed to resist Darwinism stems from our great success as creative designers. Our world is dominated by feats of engineering and works of art. We are entirely accustomed to the idea that complex elegance is an indicator of premeditated, crafted design. This is probably the most powerful reason for the belief, held by the vast majority of people that have ever lived, in some kind of supernatural deity...

Natural selection is the blind watchmaker, blind because it does not see ahead, does not plan consequences, has no purpose in view. Yet the living results of natural selection overwhelmingly impress us with the appearance of design as if by a master watchmaker, impress us with the illusion of design and planning.

(the majority of human beings apparently find it very difficult to grasp that the elegance and complexity of the natural world, as illustrated here by this manta ray, are the result of a blind process with no king, president, or CEO in charge of design and production)

I would extend the argument that Dawkins poses to include market phenomena. For many of the same reasons that Dawkins articulates, it is apparently very difficult for many/most of us to believe that a system with no one in charge of it can reliably create extremely complex, useful, or beautiful economic goods and services. Instead, we feel that any such system simply must have a Big Man, a maestro, in charge. The center of gravity of the entire discussion moves away from cultivating and protecting those decentralized processes in which individual entrepreneurs---what development economist Bill Easterly calls "searchers"---respond to the market's fitness function demand by looking for goods and services that can be sold to people via voluntary exchange. Instead, we try, again and again, to experiment with different ways to find the best Big Man (heredity? Military conquest? Popular election?).

A spontaneous order is a system which has developed not through the central direction or patronage of one or a few individuals but through the unintended consequences of the decisions of myriad individuals each pursuing their own interests through voluntary exchange, cooperation, and trial and error. This process of spontaneous evolution is not restricted to explaining the growth of the economic order but can also account for the development of language, money, culture, law, social conventions and even morals and ethics. Although the spontaneous order develops through individuals pursuing their own interest, the individuals still behave by following commonly held rules rather than by acting in a random fashion, and these rules are themselves the product of evolution.
-F.A. Hayek

Economy as Ecosystem: Planners vs. Searchers

Markets are able to operate without a grand designer because individual, independent agents, all operating according to relatively simple approach/withdraw rules, are collectively able to generate large-scale behaviors and structures that are so perfectly adapted to their environments that we have difficulty believing they could have been produced by anything but an omniscient deity.

Complexity researchers have built many models revealing how this behavior---this self-organized structure and intelligence--emerges from individual agents. In fact, very simple rules, iterated many times, can even produce universal Turing machines---essentially computers that can be programmed to perform virtually any kind of calculation.

Research into this almost eerie property of complex systems has been spearheaded by a British physicist and mathematician named Stephen Wolfram. Wolfram had been working with various simulation programs and was operating under the intuitive assumption that computerized agents following simple rules (simple "cellular automata") would generate correspondingly simple behavior patterns. He found, surprisingly, that the behavior which resulted could be extremely complex and sophisticated.

(Stephen Wolfram)

The Problem of Forecasting: The Planners are Shooting Blind

In contrast to market-based searching, which represents "the blind watchmaker" of Richard Dawkins, Big Man economies require planning, and planning in turn requires that the planner have the ability to forecast the long-term effects of political and economic programs. The central tenet of the Appeal to Intuition is that we should collectivize our wealth and then give control of it to a central planner, who can then weigh various courses of action and predict the results. It would be one thing if this were verifiably true, and political candidates were forced to present their economic forecasting models and have them independently tested for robustness and accuracy. At least then we would have some reason to believe that there could be a genuine advantage to placing resources under centralized control. The problem is that there is startling evidence that political and economic forecasting is essentially a kind of con game.

To cite just one study, Phil Tetlock of Berkeley spent some three decades rounding up the opinions of thousands of experts and then seeing how they held up to real-world events. His exhaustive study was published a few years ago and compared human experts to statistical algorithms and chimpanzees. Tetlock summarized his findings this way:

Radical skeptics...should welcome the results. Humanity barely bests the chimp, losing on one key variable and winning on the other. ...These results plunk human forecasters into an unflattering spot on the performance continuum, distressingly closer to the chimp than to the formal statistical models. Moreover, the results cannot be dismissed as aberrations... Surveying these scores across regions, time periods, and outcome variables, we find support for one of the strongest debunking predictions: it is impossible to find any domain in which humans clearly outperformed crude extrapolation algorithms, less still sophisticated ones.

(statistically speaking, we may as well put these two guys in charge)

I must return to Tetlock's work later, but in the context of mechanical trading system performance ("quant" or "black box" trading) when pitted against human discretionary trading. For now, I will simply add that the very unsettling situation that Tetlock describes is made even more sinister when he looks at how modern media campaigns work and the types of individuals who are well-equipped to get elected, despite the extraordinary evidence against the accuracy of their policy platforms and forecasts. Tetlock "notes a perversely inverse relationship between the best scientific indicators of good judgment and the qualities that the media most prizes in pundits---the single-minded determination required to prevail in ideological combat."

To recap:

1. Central planning requires that planners be able to forecast what the results of their plans will be.

2. The evidence is that this forecasting ability does not exist.

3. Those who acknowledge this, i.e., those who are cognitively best equipped to actually deal with decision-making under conditions of uncertainty (individuals who are exceptionally humble by temperament and heavily trained and oriented towards Bayesian techniques), will be unable to prevail in our soundbite world of politics, confident rhetoric, detailed "plans" and visions, and forceful campaign promises.

(In another blog entry, I will describe a spectacular problem that arises when conditions 1-3 are coupled with our old friend the martingale betting progression. Bayesians and mathematical logicians call this the "diachronic Dutch book" problem or the "money-pump" problem, and it is both amusing and tragic when applied, as it almost inevitably must be, to Big Man-style economic planning efforts).

As always, Hayek explains the situation quite beautifully:

The interaction of individuals, possessing different knowledge and different views, is what constitutes the life of thought. The growth of reason is a social process based on the existence of such differences. It is of the essence that its results cannot be predicted, that we cannot know which views will assist this growth and which will not---in short, that this growth cannot be governed by any views which we now possess without at the same time limiting it. To "plan" or "organize" the growth of mind, or for that matter, progress in general, is a contradiction in terms. The tragedy of collectivist thought is that, while it starts out to make reason supreme, it ends by destroying reason because it misconceives the process on which the growth of reason depends. Individualism is thus an attitude of humility before this social process and of tolerance to other opinions and is the exact opposite of that intellectual hubris which is at the root of the demand for comprehensive direction of the social process.

The Problem of Bundling: Probability and Complexity Favor Searching and Disfavor Central Planning

Another reason why the decentralized, distributed intelligence of the market is able to outperform an economy planned by a small cabal of elite political operators can be found in the issue of bundling. Bundling occurs when a complicated economic program with many interdependent parts is designed all at once. In contrast, the free market's idea of a rather a holistic, comprehensive program is actually the result of many smaller-scale, independent searches that are aggregated together by the emergent properties of a complex adaptive system (thereby creating a program that has the appearance of having been intelligently designed from the beginning, although of course it was not).

The mathematical support for this observation is fairly straightforward, and I'll use a format that I got from Bill Easterly: let's say that we have a bundled economic planning package, that the number of components that must work together is 20, and that the probability of success of each component is---and we are being very kind here---.85. The probability of the entire planned program being successful is:

p(program success) = p(individual component)^n ,where n=the number of components

Working through the equation we discover to our sadness that the probability of the entire program being successful is only .04, despite the high probabilities of success for each individual component. This result stems from an attempt to engineer a social system (an economy) as one would engineer a physical system like a bridge.

In contrast to the planned system, imagine an approach that is characterized by searchers who simply keep trying different things at the grassroots level until they find something that works. There is no plan; just a trial-and-error attempt to satisfy a fitness function. Let's be very hard on our searchers and suggest that each attempt made has a probability of only .2 of being successful.

The search equation: p(success) = 1 - (1-probability of attempt success)^m, where m=the number of trials.

It takes 19 experiments for the searchers to have a probability of .99 for finding something that works.

The Endgame of Big Man Economics: The Price System is Destroyed

Once the ability of a free market-determined clearing price to equilibrate supply and demand has been removed, the true costs of political initiatives can be concealed from the public. I believe that this special fiscal and monetary stealth capability is another enabling feature of the Appeal to Intuition, because it allows the planners to escape market discipline for extended periods of time and it decouples the cause-effect linkages that would allow us to actually see the patterns of value-destruction in the Big Man legacy.


The net result is that programs whose costs exceed their benefits may not only continue but expand, due to different costs of knowledge between the created constituency and the general public...(this) differential is exploited in various ways. One is the "entering wedge" approach to political innovation, in which the initial stakes are so low as to cause opposition fears to seem so exaggerated as to be discredited as outlandish. Later, the scope of the innovation can manifest itself in growing sums of money and/or burgeoning powers, after public interest has waned or turned to other things...The income tax began in 1913 with a maximum tax rate of 6 percent (!) on incomes of a million dollars per year and higher...

Temporary concealment (of true costs) pays big dividends because of the high cost to the public of trying to monitor all outgoing programs. As one federal official said (in justification), "if you put these huge capital contributions up front there's no way any administration would propose it or any Congress would approve it." In other words, the voters would never go for it if they knew.

Many economic devices and accounting tricks which do nothing more than postpone the transmission of financial knowledge to the public depend for their political effectiveness on knowledge cost differentials between the public and "insiders." One such device is simply mislabeling as "loans" expenditures which no one expects to be repaid. ...Even better for concealment purposes are "loan guarantees"...everyone involved may know that there is no rational hope that the private loans will ever be repaid, and that the banks will collect from the U.S. Treasury, eventually. In the meantime, it is not carried on the books as an expenditure or a liability (economically or politically) of the incumbent administration.

The Big Six: Archetypal Attacks on Free Markets

I think that I will conclude today's rant by briefly mentioning the six most common attacks on free market economic policy that I am personally aware of. I'll try to contend with the first one today and leave the rest for next week, because I'm getting tired and depressed and I'm sure that at this point you are, too.

1. Democracy > Markets. This idea rests on the notion that outcomes from collective action are superior to outcomes from free markets, provided that the collective action is the result of a democratic election process.

2. Behavioral Economics > TP Equilibrium Model. "TP" in this case means "Tight Prior" Equilibrium, a particular position taken by the Chicago economists that emphasized the beauty and stability of the market. This argument is sort of a take on the "God of the Gaps" gimmick that we discussed previously; it suggests that a failure of the market to live up to a standard of perfection---to have booms and busts and accounting scandals, for example---is indicative of a need for government intervention.

3. Government Provides Social Justice. This argument suggests that the outcomes from market transactions lead to an unfair distribution of wealth, so the solution is apparently to create a mechanism for generating an unfair distribution of political power, too, but then to hold the winners accountable for restoring social justice.

4. "European Socialism Works Great". The argument here is that the economic success of tax-and-regulation-heavy European welfare states shows us that free markets are overrated and that plans---at least European plans---work very well.

5. Markets Have Externality Problems. This is really the classic---usually due to high transaction costs, markets are found to have positive spillover effects (positive externalities) that should be subsidized, and negative spillover effects (negative externalities) that should be taxed, regulated, or prohibited.

6. Hope Springs Eternal. This is sort of the final, desperate line of defense for the Big Man economics disciple. It recognizes that history is replete with examples of corrupt and/or incompetent politicians making decisions that ruined economies and exacted a huge toll in terms of human misery. However, the optimist truly believes that the new guy, Candidate X, is different and that he or she will plan a better economy than we would get from a free market.

These all can made into convincing arguments and they have some good points. The six attacks that are mentioned are not meant to represent an exhaustive list of anti-market attacks, of course; they just reflect my personal experiences with those people who, perhaps even after being exposed to the case for free markets on pure performance grounds, still feel that a Big Man must be placed in charge.

1. Democracy > Markets

Those who favor majority-rule political outcomes over free market outcomes are, to me, quite interesting, because (assuming that this is not just a cynical class-warfare-based argument) their position requires that an individual voter behave ethically and/or rationally in the political context, but the same voter not behave ethically and/or rationally when buying and selling in the individual market exchange context (i.e., when "voting" with his or her pocketbook).

On the surface, it would seem that democracy and free markets are synonymous, particularly for Americans because we are generally and happily accustomed to having the two packaged together as a generalized sense of individual freedom. However, conflating the two entities can be a big mistake. The essence of the Democracy > Markets argument, which celebrates politics and the act of voting and idealizes those governments that result from popular campaigning, is that democracy provides the truly important freedom and that individual economic freedoms are somehow less important.

If a tyrant has seized power and people are not able to vote him out of office for bad behavior, then the democracy celebrant tends to admit that this would most probably lead to poor socioeconomic results for the larger population. As long as people are able to express themselves through the ballot box, however, the argument maintains that any resource-allocation regime that the democratically-elected official pursues---assuming it is consistent with his campaign platform---represents the "will of the people". Democratic political outcomes thus are said to maximize social welfare, and any economic plans that result from democratic politics should be prioritized over outcomes that a cooperative free market system of individual human actions has produced.

I suppose there is an intuitive appeal to this line of reasoning because it superficially strikes us as fair (since it promises to empower people who may not have as much power in a market setting), but the intuition is entirely misleading. First of all, it ignores the problem of a tyrant being democratically elected, which of course has happened many times, and which only needs to happen once for some of the most hideous atrocities in human history to occur (when markets fail, we get Enron; when governments fail, we get Auschwitz). If we require a planner that has a genuinely high level of competence in policy forecasting and strategic planning, then the argument essentially assumes that democratic campaigning and voting mechanisms can assure us that the rare-as-a-unicorn, market-beating, incorruptible, Bayesian intellectual demon with the 180 IQ and the Captain America-esque training in military strategy, game theory, and complexity economics will be identified and elected. This assumption is not supportable for reasons that were previously discussed in the Tetlock section.

Additionally, problems related to information asymmetries, huge cost overruns, regulatory creep, corruption and rent-seeking behaviors, and special interest groups managing to capture regulatory agencies are all relegated to a peripheral status, when in fact they act in concert to practically consume or at least contaminate all available resources within reach of their greedy, grasping hands.

In order to provide a valid economic basis for planning, the Democracy > Markets argument assumes that the net present value of various government proposals are: A) known to the policy-makers who are proposing them (this is a heroic assumption); B) accurately conveyed to voters (an equally improbable case); and C) linked to some kind of disciplined feedback system that can kill projects that turn out to have had inaccurate performance projections (now we are just kidding ourselves). Finally, it assumes that the fitness function that resource allocation searches by the voting public will attempt to meet can be clearly specified and quantified.

Beyond these, however, there is a deeper, fatal flaw in that the Democracy>Markets argument requires for the costs of voting to be high enough that aggregations of voting behaviors somehow produce more efficient results than are produced by aggregations of consumer behaviors.

In other words, if it costs me $5 to get a jar of peanut butter and I want 1,000 jars, then a free market transaction would require me to obtain $5,000 in peanut butter financing. $5,000 represents a substantive opportunity cost for me---there are many other things that I could do with that money, so we can assume that I have given the peanut butter need some thought. However, let's say that a political candidate offers me 1,000 jars of peanut butter that he will provide by taking the money from a third party, who of course would just have wasted it on trivial things, and using the money to buy me what I desire. In exchange, I must "pay" for the transaction with my vote.

If going to the ballot to vote myself the peanut butter costs me far less than the equivalent I would need to pay on a free market, then my decision-making process will change and I will believe that the true costs of goods and services converge on the marginal cost of my voting, not the market costs of the goods and services.

The result here is really no different from Sowell's discovery (discussed in the last blog post) that local zoning boards had been captured by the existing homeowners in order take land off of the market and to push up housing prices in their areas, to the detriment of those outside the system. Unless the outcomes are pre-constrained in order to protect all parties from predation, a democratic election can easily become a popularity contest for regulatory capture----the winning team gets control of the zoning board and the members of the losing team are those outside the system. Edmund Burke warned us of this danger long ago, when populist zeal and the Reign of Terror in revolutionary France made him observe that no factions in society "should be brought to regard any of the others as their proper prey."

In the peanut butter case, I haven't optimized some mythical fitness function for society, I have just found a way to use force---politically authorized force---to obtain a large peanut butter shipment for far less cost than I would have to pay on the open market. There are winners in this game (me, the peanut butter producer, the government official who got elected with the help of my vote) and there are losers in this game (the person who will now be forced to pay for my peanut butter, the goods/service provider who will now be unable to sell something, the politico who lost the election because he failed to pass my "what's in it for me?" test).

The great casualty, though, is voluntary exchange---the granular level of cooperation, the relationship between buyer and seller in an honest trade, has been diminished, and the market pricing system's ability to coordinate supply, demand, time-preferences, risk preferences, and interest rates has been wrecked. Far from being a method for exploring the fitness landscape for the greatest group utility, whatever that means, the democratic political process has in this example simply become a cheaper, expedient way for me to try to forcibly obtain things that I personally want, using a surrogate to supply the force. The difference between my vote cost and the market costs of these things will be paid for by someone else, preferably someone who isn't even born yet although I don't really care.

This is where the true Hobbesian "war of all against all" may find its most pure manifestation, and the outcomes---like nasty divorce proceedings---most benefit a third party who encourages the coercion and acrimony since it makes the party's own role even more important. For democracy to avoid these outcomes, predation by mob rule cannot be a choice; political power must be limited a priori to specific functions.

I will also note at this point that a potentially perverse result of technologically-sophisticated "direct democracy" proposals that would make voting more convenient for those who are properly equipped (say, by having voters use a secure website and identification technology) is that the cost disparity could tilt even more towards a coercive procurement method being favored over a cooperative one. On the other hand, I would say that an advantage of internet-based voting could be the ability to reduce logrolling, porkbarrel add-ons, obfuscatory bundling, and special interest legislation by having major legislative programs all subject to referendum.

Another major problem can arise if we are forced to choose between candidate-based democracy and issue-based democracy: for example, the members of Congress who had voted on the Troubled Asset Relief Program had been elected by democratic process, but the TARP---which passed through Congress---would have almost certainly failed to survive a democratic process that allowed the public to vote on that particular issue. The advantage to the political operator of retaining control and avoiding the disintermediation of a referendum is that control of legislation allows for his continued personal enjoyment of log-rolling and pork-barrel spending.

The Nail in the Coffin: Arrow's Impossibility Theorem

(Kenneth Arrow)

I turn now to what many consider to be among the most unsettling findings in all of game theory and social choice theory. Kenneth Arrow received the Noble Prize in 1972 for his research on democratic voting mechanisms. He set a few basic conditions that any voting system must meet (they are extremely reasonable and intuitive):

1. Transivity. If voters prefer Candidate A to Candidate B, and Candidate B to Candidate C, they must also prefer Candidate A to Candidate C.

2. Independence from Irrelevant Alternatives (IIA). If voters prefer Candidate A to Candidate B, and then Candidate C enters the race, they must still prefer Candidate A to Candidate B.

3. Unanimity. If every voter prefers Candidate A to Candidate B, the voting system must rank Candidate A higher than Candidate B.

4. The voting system cannot be a dictatorship.

Arrow's famous "Impossibility Theorem" reveals that no voting system we can design is able to satisfy all four requirements. Bizarre and unintended outcomes can and do emerge very easily, particularly in plurality voting systems that have more than two candidates (in the book Gaming the Vote, William Poundstone finds that at least five U.S. Presidental elections since 1828 have been determined by the presence of "spoilers", which could be said to mean that the wrong man has been elected president over 10% of the time).

Various proposals have been put forth to attempt to mitigate the effects of the Impossibility Theorem and to improve democratic elections, but they will require massive political will to implement and some pro-democracy advocates have stated, ironically, that voters are too stupid to understand them.

However, all is not lost. As we will discover, the adverse effects can be minimized if political decision-making is localized rather than concentrated, and if strict limits are placed on what can even be put up for grabs. If a democratic political process is being employed to control predatory behavior, coercion, and corruption as a "negative" or policing factor to limit government abuses, then we can in fact retain a nice linkage between democracy and market clearing prices that reflect true costs. As the political system of the United States was set up specifically as a counter-tyranny mechanism, this appears to have been the original intent of the adoption of democratically elected leaders, anyway (of course, the ballot was meant to be ultimately backed by the musket and rifle).

If, however, a political process is being used to make resource allocation decisions that are substantially different from what voluntary exchanges on a free market would have made, then democracy can become a "beggar thy neighbor" game in which individuals and groups simply seek to control the organization that has the monopoly on coercive violence (government)in order to gain advantages for themselves that they cannot gain through cooperative activity at a lower level of social granularity. As Hayek put it, "If we wish to preserve a free society, it is essential that we recognize that the desirability of a particular object is not sufficient justification for the use of coercion."

This probably seems obvious to many readers, so I will simply summarize with this graph (taken from a study that was commissioned by the Congressional Joint Economic Committee during the Clinton Administration), which shows the inverse relationship between GDP growth and government size (a function of government spending as a percentage of GDP).

Someone will no doubt bring up the "correlation is not causality" issue, which of course is true, so I will add that it could certainly be the case that GDP growth and limited government size are both "caused" by a third variable that is not present in the chart (i.e., an orientation towards free markets could cause both GDP growth to be high and government size to be low). However, this really would only serve to emphasize the same point.

To be fair, I must state that I have heard another, extremely cynical argument for favoring democratic political outcomes over market outcomes, and this went something like this: Sure, markets are great and ideally we would have extremely limited government. However, people are inherently angry and jealous and need to be able to feel that they can flex political muscle, and to be entertained with bread and circuses every now and then. Thus, if we didn't have a "pressure relief-valve" at the ballot, wealth would become too concentrated for the majority's tastes and the whole capitalist system would be brought down by violent protests, vandalism, and thuggery.

I myself am skeptical about this fatalistic argument, but I'll leave the reader to ponder it. It is an old one, it sometimes seemed to be showing up in Schumpeter's works, and it has been brought up many times to support FDR's "New Deal" programs during the Great Depression (including by the late Robert McNamara).


With thanks to my colleagues Sinnerman Mike and Jay Swan for finding it, I close with this clever macroeconomic hip-hop fable:


  1. Great post... I've always thought there had to be some evolutionary tie-in to the urge for top-down planning, but I'd never investigated it in any systematic way. Thanks!

    It seems to me that you're using the terms "planning", "regulation", and "intervention" somewhat interchangeably. Is this intentional? Bear in mind that I have very little background in this area, but it's not obvious to me that they're even close to the same thing, nor that the arguments against them are.

  2. This comment has been removed by the author.

  3. That's a great point re: dangers of assuming interchangeable terms. I need one umbrella term for "government price manipulation" in its various forms.

    I am really talking about a third party attempting to set prices independently of the market mechanism, either through forbidding a transaction which would otherwise have taken place between a willing buyer and seller, or through actually setting the price by fiat.

    The intention to influence prices may be indirect---subsidizing this, taxing that, bailing out this, creating higher compliance costs (through regulatory agencies) for that.

    I should really be more careful because some view "regulation" as a benign policing-type activity and assume that it is basically price-neutral outside of criminal behavior (just a way to manage those negative externalities), and only consider the most invasive forms of government intervention---outright nationalization of an entire industry, say---as representing "economic central planning." I have a much lower standard and simply mean that a voluntary transaction between buyer and seller would have cleared the market at a certain price, but that price/transaction is not occurring because the government got involved.

    Maybe "third-party price distortion" is a better catch-all term...? I need a phrase that implies general interference. Any thoughts on terminology from a Stanford-educated man of letters?

  4. Blogger comments seem to be completely broken for Firefox on Ubuntu, so here's hoping that Opera works... I'm tired of browser failures...

    I think "third-party price distortion" is a great general term, although it's probably best to be as specific as possible.

    The reason I asked about this is that like most disciplines, economics and finance have a lot of terms that seem general to a layman like me, but actually have a specific technical meaning to someone like you--I wanted to make sure that this wasn't one of those cases.

    The whole concept of a "voluntary transaction between a buyer and a seller" seems like another possible expression of this kind: to me it seems fraught with peril due to the wiggle room present in each of those terms ("voluntary", "transaction", etc), but I suspect that you mean something quite specific by it.

  5. I am pleased for you to use my copyrighted manta ray image but could you please e-mail me at confirming that you acknowledge my copyright and also place this following text
    © next to my image. Thank you. Malcolm