Thursday, August 16, 2012

Does Oxytocin Solve the Adam Smith Problem? Part 2

In my previous post, I summarized Paul Zak's argument for oxytocin as the "Moral Molecule."  Now I turn to the five difficulties that I see in his reasoning.

The most fundamental difficulty is that some of Zak's conclusions are not strongly supported by the evidence he presents. 

For instance, one of his most important conclusions is that in his experiments with the Trust Game, when player-B receives an intentional signal of trust from player-A, because player-A has shared some money, there is a strong correlation between player-B's level of oxytocin and this player's trustworthiness in returning some of the money to player-A.  This correlation shows up for most individuals, but not all.  Some individuals acting as player-B receive intentional signals of trust and show a surge of oxytocin, but they return none or very little of the money they have to player-A.  The technical term for such people is unconditional nonreciprocators.  Zak calls them bastards.

Can Zak explain bastards in a way that supports his conclusions?

When this question comes up in Chapter 5 ("The Disconnected") of The Moral Molecule, Zak cites only one of his experimental reports--an article published in 2005 in Behavioral and Brain Sciences.  In that article, the only presentation of data is a scatter diagram, which shows a scatter of points between levels of oxytocin in the blood and trustworthy behavior (dollars returned by player-Bs to player-As after the tripled transfer from player-A to player-B) for player-Bs receiving an intentional signal of trust.  In that article, Zak claims that this scatter shows a "positive relationship" between levels of oxytocin and the trustworthy behavior of player-Bs (Zak 2005, 369).  In fact, as John Conlisk (2011) has pointed out, there is no statistically significant "positive relationship" in the scatter, unless one throws out the five cases of individuals who showed high levels of oxytocin while refusing to show  trustworthy reciprocation.

To justify throwing out these five individuals, Zak explained:
We investigated traits that differentiated these five 'usual' participants from the others and found that they exhibited labile affect on four self-report measures, were usually sexually active, said that they thought others were trustworthy and evaluated themselves as very trustworthy.  They also stated that accumulating wealth while others lived in poverty was acceptable.  Though these results are based on a small sample and should be taken with caution, they suggest that a lack of trustworthiness after receiving a signal of trust is associated with identifiable personality traits. (Zak 2005, 369)
Oddly, Zak does not identify these "identifiable personality traits."  Being sexually active?  Thinking that others are trustworthy?  Thinking that it is acceptable to accumulate wealth while others live in poverty?

In 2008, Zak returned to this point in a chapter for his edited book Moral Markets.  He explained:
. . . I studied 212 subjects making trusting decisions and showed that their brains released oxytocin approximately proportional to the intentional monetary signal of trust received from a stranger: the stronger the signal of trust, the more oxytocin is released.  Approximately 98 percent of thse subjects also had proportional behavioral responses: the higher their oxytocin levels, the more they shared money with the person who initially demonstrated trust in them.  But the other 2 percent of subjects, though their brains produced a surge of oxytocin, were untrustworthy, keeping all or nearly all the pot of money they controlled (Zak, 2005).  Two percent is roughly the proportion of sociopathy in the population, and these subjects psychological profiles had elements of sociopathy.  A discussion of the neural mechanisms that produced this behavior is beyond the scope of this chapter . . . (Zak 2008, 268)
So now the bastards--those individuals who do not show the predicted correlation between surges of oxytocin and trustworthy behavior--are to be thrown out because they are sociopaths.  But it is not clear how he justifies identifying them as sociopaths.  Is it because we know that 2% of any random group of people are sociopaths?  How exactly did their "psychological profiles" show sociopathy? 

These individuals were untrustworthy even though their brains showed "a surge of oxytocin."  Does that mean that surging oxytocin by itself is not necessarily correlated with trustworthy behavior?  If so, does that weaken Zak's whole argument for oxytocin as "The Moral Molecule"?

In his new book, Zak returns again to this point.  Now he reports that the proportion of bastards has increased from 2% to 5% of the college students in his experiments.  He identifies this 5% as suffering from "Oxytocin Deficit Disorder."  But he admits that this might seem odd since they show "an excess of oxytocin" (104).  He explains:
. . . the system does not react to the overall level of oxytocin but only to the immediate surge.  The off-switch in their receptors was malfunctioning, flooding the system with oxytocin, and the overflow was creating a functional deficit.  No surge, no contrast, no oxytocin activation.  No oxytocin activation, no empathy or reciprocity. . . .
There are two problems with this explanation.  First, he does not explain how he knows that "the off-switch in their receptors was malfunctioning."  Did he determine this through a medical examination?  Or is this just speculation?

The second problem is that while here in the new book he says there was "no surge" of oxytocin in these individuals, he previously reported (in the quotation above) that "their brains produced a surge of oxytocin."

So there is some confusion in Zak's accounts of his experimental results.  Conlisk reports that when he asked to see Zak's data for these experiments, he was told that the "data had been discarded within old computers" (Conlisk 2011, 161).

Perhaps the only firm conclusion we can draw from this is that there's no accounting for bastards!

To assert that oxytocin is the molecule for love and morality is a bold claim.  To render such a bold claim plausible, one would have to support it with deep research to uncover exactly how oxytocin functions in the complex neurophysiological networks of the human brain and body.  Zak has not done that.  And he cannot do it with the methods he employs.

Zak says that his Dean at Claremont Graduate University once described his research as "vampire economics."  It's a good label, because Zak's method is to have individuals play behavioral economic games and then to take blood samples from them and measure the level of oxytocin to look for correlations of the oxytocin with their behavior.  Measuring oxytocin in blood samples is a shallow probe into the complex reality of the neuroendrocrine systems of the brain and body.  It's too shallow to reveal the complex mechanisms by which oxytocin might influence the complex behaviors of love and morality.

Some of the leading researchers in biological psychology deny that the evidence of neuroscience supports any identification of oxytocin as the Love Molecule or the Moral Molecule (see, for example, Panksepp and Biven, 2012, 38-42).  How would Zak resolve this dispute in his favor?  He can't do that because this would require research that would probe much more deeply than is possible with the shallow methods of vampire economics.

The crucial part of Zak's argument for moral markets is the "Oxytocin Prosperity Cycle":  oxytocin promotes empathy, which promotes morality, which promotes trust, which promotes prosperity, which promotes even more oxytocin.

One important line of reasoning for this conclusion depends on Zak's analysis of data from the "World Values Survey" showing a correlation between people reporting attitudes of tolerance and trust and the average income in their nations, so that national tolerance and trust correlate with national wealth (Zak 2012, 171-73).

There are three problems with this analysis.  The first is that Zak does not have any international data on oxytocin to indicate that countries with high values for trust and tolerance also have high surges of oxytocin.

The second problem is that high national trust might be best explained as a result of high ethnic similarity.  Of the four nations reporting the highest levels of trust, three are Scandinavian--Norway, Sweden, and Finland--nations known for their ethnic homogeneity.  In fact, Zak recognizes this: "countries whose citizens are more similar have higher trust--the cognitive and affective mechanisms that induce the understanding of another's intentions may simply be easier to read when those around us are similar" (2008, 273).  Of course, Zak might claim that this still shows the work of oxytocin, because oxytocin surges more easily in social interactions between people who are similar to one another.

The third problem is that Zak reports that from 1960 to the present, the United States has shown a dramatic drop in the proportion of Americans who trust one another (2012, 177).  Since this same period showed the greatest increase in national wealth in American history, it seems that levels of trust and levels of wealth have been inversely related, which is the opposite of what Zak would predict.

Sometimes, Zak offers a strong version of his main thesis:  "modern market exchange is inconceivable without moral values," and these moral values arise "in the normal course of human interaction, without overt enforcement" (2008, xi, xvii).  The obvious objection to this is that this is simply not true: many people in modern markets will behave immorally if good conduct is not enforced legally through clear threats of governmental punishment.

In response to this objection, Zak can move to a weaker version of his thesis: "most people, most of the time, behave ethically," but many people will behave unethically in certain circumstances--such as in the unethical environment of the Enron company, for example--and to mitigate such moral market failure, we need economies that are at least "moderately regulated" by law (2008, 261, 276).

Zak tells the story of J. Clifford Baxter, who became vice chairman of Enron in 2000.  Baxter was a moral man, and he complained about the unethical business practices of the company.  He finally resigned, and then committed suicide because he could not endure the pain of what had happened in Enron (Zak 2008, 259-60).  Zak does not bring up this case in his new book, but he should have, because it illustrates moral market failures, even as it illustrates the disgust with such moral failures felt by morally sensitive individuals like Cliff Baxter.

Moral failures like Enron become more likely to the extent that modern markets are impersonal.  The neurophysiological basis of morality studied by Zak is most effective in personal interactions.  But as modern markets become extended, and numberless individuals from around the globe engage in trade with distant strangers who might not even be known to them, it becomes ever harder for the moral sentiments to be aroused.  Zak recognizes this problem:
The move from personal exchange to modern, mostly impersonal exchange in markets is the key to the division of labor that caused the rapid gains in productivity and wealth since the Industrial Revolution (Vernon Smith, 2003).  Because the instantiation of values varies somewhat across both individuals and environments, violations of values must have consequences.  Enforcement in traditional societies is personal--you cheat me, then I hurt, or ostracize, you.  The incentives to cheat, free ride, and steal are rampant during impersonal exchange, necessitating an enforcement body that all accept, namely, government.  (2008, 275)
Like all classical liberals, Zak is suspicious of government and its "top-down" regulation of markets, because he thinks that in general the "bottom-up" emergence of unintended order in markets is more efficient and more moral.  But again like all classical liberals, he recognizes that markets always need rules, and impersonal markets need some rules of law enforced by government.  He does not stress this enough in his new book, but he does at least acknowledge the need for the "procedural fairness" that comes from government providing "equality under law, an impartial judiciary, freedom of the press and assembly, and the light to moderate economic regulation that allows the overall economic pie to expand" (2012, 200-201).  He also recognizes that to "keep the oxytocin flowing," we need a short-term "safety net" for the poorest people, and we need to avoid gross inequality of wealth.

All of this is consistent with the tradition of classical liberalism from Adam Smith to Friedrich Hayek, which has recognized the need for government to provide the minimal procedural rules that increase the probability that impersonal markets will become moral markets.

Does Zak's argument for oxytocin as the "Moral Molecule" solve the Adam Smith Problem? 

His argument does show how modern neurophysiological research can confirm the moral psychology of Smith's Theory of Moral Sentiments by showing how it can be rooted in the evolved biological psychology of the human species.

And if we read The Wealth of Nations in the light of the Theory of Moral Sentiments, we can see how Smith's economics can be embedded within his moral philosophy. 

But this does not completely resolve the Adam Smith Problem, because it does not explain why Smith wrote The Wealth of Nations without any references to his moral teaching in the Theory of Moral Sentiments.  Zak says that Smith teaches "that the pursuit of self-interest can indeed benefit all, but only so long as it takes into account the mutual sympathy that leavens the contrary forces that are almost always at work in us, namely, greed and aggression" (2012, 168).  But the very term "mutual sympathy" never appears in The Wealth of Nations!  And if we look at Smith's Index for The Wealth of Nations, we see an entry for "Self-love the governing principle in the intercourse of human society, 26-7."

If one reads The Wealth of Nations by itself, one could easily conclude that economic science can explain human motivation with nothing more than the axiom of self-interest as "the governing principle in the intercourse of human society."

Adam Smith could have avoided the Adam Smith Problem by making clear to the reader of The Wealth of Nations that the economic teaching of that book is embedded within the moral teaching of The Theory of Moral Sentiments.


Conlisk, John (2011).  "Professor Zak's Empirical Studies on Trust and Oxytocin," Journal of Economic Behavior & Organization 78: 160-66.

Panksepp, Jaak, and Lucy Biven (2012).  The Archaeology of Mind: Neuroevolutionary Origins of Human Emotions (New York: Norton).

Zak, Paul J. (2005).  "Trust: A Temporary Human Attachment Facilitated by Oxytocin," Behavioral and Brain Sciences 28: 368-69.

Zak, Paul J., ed. (2008).  Moral Markets: The Critical Role of Values in the Economy (Princeton, NJ: Princeton University Press).

Zak, Paul J. (2012).  The Moral Molecule: The Source of Love and Prosperity (New York: Dutton).

Some of my posts on related topics can be found here, here, here, here, here, here, here, here, here, and  here.

No comments: