“I think nobody can understand markets because they are totally irrational.” – Nobel laureate Professor Joseph Stiglitz, Columbia University, BBC Newshour, 8/5/11
Traditional study of economics presumes that man is rational, calculating, and acquisitive. Economists know that we are not steadfast in these dispositions, but the presumption works because our deviation from these characteristics is relatively rare. Economic modeling need not be comprehensively predictive to valuably advance the science. As a matter of fact, it would not be called ‘modeling’ if it took every last detail into account.
But over the past few decades, a new guard has emerged, referred to generally as ‘behavioral economists.’ To degrees that vary within the sub-discipline, behavioral economists believe that the traditional presumptions about ‘economic man’ and, by extension, the operations of the macroeconomic market, are invalid. According to behavioral economists, humans are largely irrational when it comes to money matters.
The financial crisis of 2008 and its aftermath have raised the profile of behavioral economics. But the seeds of behavioralism were planted during the last major financial disaster, the Great Depression, by John Maynard Keynes. He famously blamed “animal spirits” for a market that “can remain irrational a lot longer than you and I can remain solvent.”
Today’s behavioral economists are not content to cast ‘animal spirits’ as a mere aspect of our economic existence. Their goal is to put human irrationality at the very center of economic study.
The locus of work on behavioral economics has shifted from from the offices of economics professors to the laboratory; from the theoretical to the empirical. Its advancement now relies heavily on social psychologists like Professor Jennifer Lerner of Harvard, who told PBS’s Nova in April 2010 that “much of the behavior that led up to the (2008) crash is unexplained by the discipline of economics.”
The episode of Nova featuring Professor Lerner, “Mind Over Money,” served as a sort of public coming-out party for behavioral economics. The program opens with an experiment involving the auction of a $20 bill. According to the voiceover, "(t)he auction is a key experiment for behavioral economists … that challenges all our ideas about money.” It is set up like a typical auction, with about a dozen participants. But “there’s a catch … the second-highest bidder receives nothing, but pays the amount of the losing bid.”
Now, if one were to sit down in advance and really contemplate this ‘catch,’ it would become clear that the smartest move is to decline to bid in the auction at all. Under the auction rules, if I bid, say, $19, and then a fellow participant bids $20, then I am out $19 with nothing to show for it if I resign. So by that point, I will likely continue bidding upwards, even though it is apparent that I am locked in a bad situation.
But the participants in the Nova experiment were there to participate, so they jumped right in. Unsurprisingly, a couple of them realized the implications of the ‘catch’ only after it was too late. “Amazingly, two participants bid way above the face value … so why would anyone pay more than $20 for a $20 bill?” “It’s a trap that if you don’t think more than a couple steps ahead, you fall into … nobody will want to play that game twice,” says University of Chicago economics professor and prominent behavioralist Richard Thaler, about a rigged experiment that actually does nothing to advance his favored theory of human economic behavior.
The showcasing of this auction experiment demonstrates the broad confusion within the behavioral economics community between poor strategizing, or even stupidity, with irrationality. The two final competing bidders did not have an aberrant sense of what a $20 bill is worth, nor were they driven by some sort of frenzied desire to win, as the Nova voiceover presumptively asserts. They were simply, as Professor Thaler observes, trapped before they fully figured things out. Once they had a handle on the situation they were in, they finished out the auction in a perfectly rational manner, each hoping the other would resign before the bidding climbed into the stratosphere. (The ‘winner’ paid $28 for the $20 bill, thus incurring a net loss of only $8. His opponent was out $27.)
The Nova episode covers a couple more experiments that prove, upon examination, to be equally invalid showcases of irrational behavior.
In one, a few ‘men on the street’ are asked first if they would prefer to be given $100 a year from now, or $102 in a year and a day. All choose the $102. But when the same people are asked if they would prefer $100 now or $102 tomorrow, all choose the $100 now, despite the determination by the voiceover that “the larger amount is still the rational choice.”
Now, different people have different tolerances for deferral of gratification. In some sense, the entire capitalist ethic is based on deferred gratification: the more one invests, instead of immediately consuming, the more one can consume overall, over time. In the early 1970s, Stanford psychologist Walter Mischel tested four-year-olds’ ability to defer gratification by offering them one marshmallow now versus two in fifteen minutes. The children who held out for a second marshmallow went on to earn markedly higher incomes in later life than those who did not.
But no one would suggest that the children who gave in to their immediate marshmallow cravings behaved irrationally. In the same vein, the worst that could legitimately be said about interviewees who want $100 now instead of $102 tomorrow is that they are not prudent investors. (It is a stretch to put even that on the Nova interviewees, given that two dollars would barely buy a small cup of coffee. It would have been more revealing if the show’s producers had pitted $1,000,000 against $1,020,000…)
As for the contrast set up by these same subjects opting for $102 in a year-and-a-day, rather than $100 in a year: in that situation, they would have to wait only one quarter of one percent longer for the $102. But mathematically, tomorrow is infinitely farther away than now is.
There is a further sensibility to the apparent ‘present bias’ that the Nova interviewees display. When a random stranger says he will give you money tomorrow, instead of right now, even a figurative calculation must consider whether the money will actually be delivered tomorrow. But there is an equal chance that the stranger will come through in 366 days versus 365. That people leave any money at all in low-interest bank accounts demonstrates that there must be some reason other than present bias behind the opting for $100 today. If Nova’s producers put $100 into each of the interviewees’ bank accounts, and then reliably demonstrated that they would add two dollars if the principal were left alone for a day, would any but the most immediately desperate subjects rush to cash the $100 out? Most adult Americans can ‘have’ or consume $100 any time they want.
Nova showcases a third experiment involving the valuation of an insulated, lidded coffee mug. A group of students is asked what each of them would pay for such a mug. The average reply is six dollars. Then each student is given a mug for free, dismissed, and told to return an hour later. Upon their return, the students are asked what they would be willing to sell their mug back for. The average reply is nine dollars. The voiceover tells us that “(t)he emotional pleasure of owning something for just an hour pushed the price up by 50% … In rational economics, the price should be exactly the same. After all, the value hasn’t changed.”
Where to start with this one? First, it is quite a thing to say that “the value hasn’t changed.” Obviously, it did change! In an exchange that is largely isolated from the market, like the mug sell-back situation, value is an almost purely subjective notion. Perhaps, during the hour interim, the students got to thinking about how they would use the mug. Perhaps some did use it, and then strapped it to their backpack. After an hour, some students were likely already thinking in terms of replacement cost. If they sold the mug back, how long would it take them to track down a new one? How much would a new one cost, once located?
And maybe emotion did play some part in the students’ revaluation. (One student said that owning the mug for an hour made it seem “a little bit more special, because it was going to be something useful to me.”) Emotions themselves do not rise from strictly rational processes, but satisfaction of emotion is a standard component of rational cost/benefit analysis. We buy gifts for others because we feel affection for them. We give to charity out of compassion. We bring home the dog that captures our heart. Raising the valuation of the mug would have been irrational only if owning it had given rise to a negative emotion. Even the very basic model of ‘economic man’ does not suppose that we are Mr. Spock.
Besides, rational economics would never expect a person to sell something back at the same price they paid for it, even immediately after receipt. If I am willing to part with six dollars to obtain an item, then clearly I would rather have the item than the six dollars. So I value the item more than I value six dollars. If my valuation of the item were static, I would require something more than six dollars to relinquish ownership.
Again, none of this is to suggest that behavioral economics is complete hooey. Humans do deviate from the rational model, sometimes in notably consistent ways. Behavioral economists and their psychologist adjuncts have made a convincing case that we are loss-averse, meaning that losing a given amount pains us more than that same amount’s gain satisfies us. (However, loss-aversion is at least partially explained by the law of diminishing marginal utility.) And one experiment showcased in the Nova program aptly demonstrates ‘anchoring,’ a phenomenon familiar to merchants who prominently display a ‘regular price’ along with a (perhaps permanent) ‘sale price.’
The behavioralists’ most potent challenge to the framework of rational economics is based on the market’s periodic tendency to sharply inflate and then rapidly deflate the price of a selected class of assets, also known as the ‘bubble’ phenomenon. Given that we are still suffering from last decade’s massive housing bubble, this is a highly opportune time for advocates of behavioral economics’ primacy to use bubbles to capture the public imagination.
At first glance, bubbles seem to engender some sort of irrationality. Surely, the value of a subject asset class does not bubble up and burst, as pricing patterns would suggest.
Chapman University professor Vernon Smith pioneered a set of ‘experimental asset market’ lab studies starting in the 1970s. The studies evolved into an exploration of the psychological underpinnings of human ‘bubble behavior.’
The most basic of Professor Smith’s experiments is showcased in the Nova program. In it, test subjects trade a single stock with each other in a simulated market, through personal computer terminals. They are told beforehand that with each of the fifteen discrete trading sessions, the ‘fundamental value’ of the stock will fall, eventually hitting zero. (This decline is noted, as trading progresses, on the computer displays.) After each discrete session, the subjects’ simulated cash holdings are augmented by a dividend of randomly-determined magnitude. At the experiment’s end, the subjects receive real cash for the simulated cash and stock they are left with, which could amount to hundreds of dollars.
In this scenario, most first-time-player subject groups initially bid the price of the stock far above its fundamental value, but bid the price down to almost zero by the end of trading. In other words, they collectively exhibit classic bubble behavior. The experiment is presented by Nova as almost conclusive evidence that economic man is innately and systematically irrational.
But there are a few problems with using this experiment to support such a conclusion. First, Professor Smith himself found that by the third time a given subject group runs through the experiment, the bubble behavior disappears almost entirely. This suggests that the test subjects’ initial wildly high bidding stems largely from knowledge that they are there to play a trading game. Much like the $20 bill auction subjects, they jump right in and start playing. But once they get the hang of it, they begin to truly strategize. Nova neglects to mention this exhibition of solidly rational adjustment behavior on the part of repeat players.
Nova also neglects to acknowledge the dividend payments in their description of the experiment. Fortunately, the potential impact of dividends did not escape examination by Professor Smith himself. In 2000, he found that bubble behavior in this type of experiment is greatly moderated, if not completely eradicated, when the turn-by-turn dividend payments are eliminated and the ‘fundamental value’ indicator is held constant. As Professor Smith told the website bigthink.com in late 2009, “(l)aboratory results make it very clear that it’s cash flopping around the system that tends to give you these runaway asset market bubbles.”
In a 2010 study, three University of Innsbruck researchers replicated Professor Smith’s experiments and found, through post-session questionnaires, that “most subjects do not understand the declining fundamental value process … (r)unning the experiment with a different context (‘stocks of a depletable gold mine’ instead of ‘stocks’) significantly reduces mispricing and overvaluation…”.
That first-time subjects do not understand the ‘fundamental value’ variable should come as no surprise. After all, ‘fundamental value’ has no real-life analog in an exchange-based market. An individual’s valuation of an isolated item (like a coffee mug) is a subjective determination, but a market asset’s value is, to a great degree, a function of its market price. Price and value cannot really be decoupled, as they are in the original version of Professor Smith’s experiment. In an exchange-based market, an asset’s perceived value is tied much more closely to what others will pay for it than it is to any individual conception of its utility.
Among the behavioralist intelligentsia, Professor Smith’s experiments are often used to attack the Efficient Markets Hypothesis (EMH), which is most strongly associated with University of Chicago professor Eugene Fama. According to Nova, EMH posits that “(market asset) prices at any moment in time cannot be wrong.”
This is, at best, an inelegant description of EMH. At worst, it is a complete strawman. What could it possibly mean for a market price to be ‘wrong?’ If, on an open market, there is a willing buyer and a willing seller at a given price – which is the definition of a ‘market price,’ how can that price be ‘wrong,’ absent specific fraud? No, all EMH truly purports to describe is a state of price equilibrium, not price ‘rightness.’
In trying to correct the widespread misperceptions about EMH, Eugene Fama’s son told ifa.com (12/10/08), “(i)t actually means that everybody making multitudinous decisions … everything they know in combination is already in the prices to such an extent that no single participant … can outperform by more than you can really expect by chance and do it persistently.” EMH is, in its essence, about information.
So Professor Smith’s experiments actually mesh quite well with EMH. As subjects’ knowledge about the implications of the experiments’ ‘fundamental value’ variable increases, their bubble behavior moderates. But the constant payment of dividends increases bubble behavior, since with each trading round the subjects have new information about how much cash they have to play with. In real-world bubbles, constant cash infusions (caused often, though not exclusively, by a central bank’s loose credit policies) are ongoing new information as a bubble inflates.
But in the real world, much more so than in Professor Smith’s experiments, the greatest bubble-inflating (or deflating) new information is that prices rose (or fell) yet again yesterday. ‘Herd mentality’ is eagerly identified by behavioralists as an irrational phenomenon, but really it is just a manifestation of questionable individual decision-making. Every day of price trending is new information about others’ opinions of what a class of assets is worth. Every day added on to a trend increases the likelihood that that trend will continue tomorrow.
Of course, those who do not properly absorb the laws of probability (a form of generalized information) fail to fully grasp that a low-probability event will occur eventually, with enough iteration. These are the unfortunates who are left holding the bag when a bubble begins to deflate. (Regarding the recent housing bubble specifically, investors had the additional misleading knowledge that housing prices had never persistently fallen in modern times.)
Once upon a time, it was in vogue among rationalists to maintain that volitional market directionalities are a mere illusion, that asset prices’ chart patterns always describe a ‘random walk.’ Over the past several years, rational economics’ leading lights, including Professor Fama, have backed off from at least the absolutist version of this assertion. ‘Bubble behavior’ is real, and bubbles are generally disruptive to the smooth functioning of the economy… but there is nothing inherently irrational about them.
Behavioral economics is a worthy niche in the greater discipline of economic study. Since Adam Smith set it on its course, economic science has matured to the point where the marginalia of economic man’s doings merit attention.
It is the attempt to use such marginal behavior to overturn the rational actor model that is objectionable. We must not allow the implications of specific behavioral quirks, magnified in psychology laboratories, to be exaggerated by those seeking the professional prestige and grant money afforded to academic giant killers.
Beyond its academic practitioners, behavioral economics has found eager acceptance among advocates of an activist central government. If economic freedom abets habitual and pervasive irrational behavior on a micro level, and so leads to irrational results on a market level, then government has a duty to sling a marionette wire around the invisible hand.
But if discrete decision-makers, operating within a just legal framework, can be trusted to foster a predominantly rational market, then the case for extensive governmental interference is left to rest solely on well-worn appeals to socialist idealism.