Economics Can Explain the Real World with Two Fixes

Why rationality should be based on relative payoffs

Share with your friends

More share buttons
Share on Pinterest

By Lixing Sun

Rational choice theory lies in the heart of neoclassical economics. It assumes that humans are utility maximizers, always going for the option with the highest net benefit. Since the 1970s, however, behavioral economists have poked so many holes in the theory that pronouncing its death now feels like flogging a dead horse. Can we save rational choice theory and revitalize neoclassical economics? If so, how? Here is a 2-point proposal as an attempt to infuse evolutionary ideas into our economic thinking.

“Rational” Is Relative

Consider revenge and retaliation. Why would we often engage in such spiteful actions ruinous for all parties involved? An early clue emerged in 1960 from a little-known economic experiment, in which college students were asked to play a repeated two-person, non-zero sum game. In the game, two players separated by an opaque screen can push two buttons, one for “Cooperate” and the other for “Defect,” with the following payoffs:

Table 1. A “Cooperation Game” with absolute payoffs

Screen Shot 2016-04-12 at 11.24.49 PM

Given the payoff matrix in Table 1, what will the players do? The answer based on neoclassical economics is that they both opt for “Cooperate” because this will yield the highest profit for both (4¢ per round). Since mutual cooperation brings in a better return than any other strategic combination for both players, it appears unbeatable—a Nash equilibrium in the lingo of game theory. The game is accordingly called a Cooperation Game. Here, the paired numbers in each of the four cells are the payoffs for Player 1 and Player 2, respectively, corresponding to the strategies (“Cooperate” or “Defect”) they choose. For instance, if both players choose “Cooperate,” their payoffs are the same, 4¢, or (4¢, 4¢) in the top left cell in Table 1. (Note: 1¢ in 1960 is worth 8¢ today.)

Get Evonomics in your inbox

When real people played the game in the study, however, the result were anything but economically rational: nearly half of the players chose to defect! How can Homo sapiens stray so far from Homo economicus?

The answer lies in a hidden glitch in our reasoning process. In this economic game, what is considered “rational” in neoclassical economics is absolute payoff. The real gauge in evolution, however, is relative payoff. The difference becomes obvious if we convert absolute payoffs in Table 1 into relative payoffs—the net differences between the two players for the four payoff cells—in Table 2:

Table 2. The same “Cooperation Game” with relative payoffs

Screen Shot 2016-04-12 at 11.26.31 PM

Why are we prone to playing the Spite Game? What will you play now? Table 2 shows that the best possible relative payoff for you is the same for either strategy, but if you play “Cooperate,” you risk being suckered when your partner plays “Defect”: -2¢ for you and 0¢ for your partner. So, playing “Defect” guarantees the best possible relative payoff, regardless of what your partner does. Since your partner is also a thinking Homo sapiens, you and your partner become locked in mutual defection. (Technically, “Cooperate” is weakly dominated by “Defect” for both players in this game.) As such, what begins as a seemingly Cooperation Game in Table 1—where absolute payoff is what counts—reveals its true color of what I call a Spite Game in Table 2, where it is the relative payoff that matters a great deal. (See economist Thomas Riechmann’s paper for a general treatment about this issue.)

Apparently, according to biologist David Barash’s book, The Survival Game, we often attempt to maximize the difference in payoff among peers, as though we compete to gain the upper hand in a small tribe—a milieu in which the human brain evolved. Thus, as it seems, spite isn’t evolution’s sloppy job in designing human mentality. Rather, it’s a logical consequence of evolution operating according to relative payoff.

Naturally, spite can’t be a curse for humans only. Similar behaviors have also been observed in other organisms. Bacteria, for example, manufacture toxins that kill their conspecific neighbors. Sticklebacks raid the nests and munch on the eggs of their own species. Swallows defend large territories to jeopardize the reproductive chances of their competitors. Monkeys harass mating couples in their cohort. Perhaps, that’s why humans have evolved the habit of making love in private.

Like many other organisms, we humans can gain a competitive edge by either doing better than our rivals or making our rivals do worse than us, regardless of whether both parties are better off or worse off in the absolute sense. Putting it differently, competition is about gaining relative payoff—or, relative utility in economic terms.

Clearly, one way to infuse Homo economicus with the blood of Homo sapiens is to enrich the meaning of “rational,” which should be gauged by relative utility, not absolute utility, especially when we deal with rivalries in direct competition. This will make otherwise “irrational” choices—envy, revenge, retaliation—completely rational.

Of course, relativity in payoff isn’t the entire story. Rationality also depends on how we use the time frame.

“Rational” Is a Lifetime Deal

Consider an economic Ultimatum Game, where two people, a Proposer and a Responder, are given some money, say $10 or $100. The Proposer suggests a way to split the money between them. If the Responder agrees, each keeps the proportion of the money as proposed. If the Responder rejects it, neither of them gets anything.

According to rational choice theory, the Proposer should give the Responder as little as possible, and the Responder should accept any non-zero offer from the Proposer. Experimental results from real people in industrial societies, summarized in economist Colin Camerer’s book, Behavioral Game Theory, however, are starkly different: the Proposer’s offer is typically above 30%, and the Responder rejects the Proposer’s offer about half the time when the offer falls below 20%.

Why, again, are Homo sapiens so unlike Homo economicus?

The key to the answer lies in whether we see the issue in the short or long term, in addition to using the concept of relative payoff.

The disparity between reality and rational choice theory stems mainly from the fact that most Ultimatum Game experiments are set up as one-shot deals between strangers. This setting is quite detached from the historical milieu in which our brains evolved. Despite radical cultural changes, much of our instinctual behavior today is still tuned to life in the Stone Age, when our ancestors were living in small, close-knit bands. In this ancestral environment, if you bumped into another person, it was unlikely that you would never meet him or her again. As such, our brains are largely tuned to repeated social dealings in the long run. This is not what most game theory experimenters presume in their studies.

How might a mentality tuned to such a long-term vision be favored by natural selection? One factor is that the Proposer’s generality can encourage the Responder to return the favor in the future. Such a reciprocal partnership can benefit both parties over time.

Another factor favoring a long-term perspective is reputation. A good reputation endows a halo of trustworthiness—a badge for being cooperative—to the Proposer. A person with a good reputation, as shown by mathematicians Martin Nowak and Karl Sigmund, often gets help from people for free—free in terms of not having to return the favor. As such, a good reputation earns one long-term profit in a stable community, a fact well reflected in folk wisdom in many societies—“Kindness will pay off.”

The value of reputation can explain such enigmatic human customs as dueling in Europe, “saving face” in East and South Asia, and honor killing in some Muslim communities. For instance, the ancient Chinese practiced filial piety, in which family elders were revered to a degree that they were entitled to the best living conditions—the finest food, the coziest rooms, and the highest social status—even though they were too old to work. The logic: filial piety served as the gold standard for reputation. (How can we trust a person who won’t even treat his parents or grandparents well?) Without knowing the value of reputation, the practice of filial piety makes no economic sense. Clearly, the benefit of reputation warranted the cost of sacrifice.

Today, even without conscious learning, our genes, through their expression as our instincts and intuitions, still send the old, stubborn instructions to our minds, making us behave as if we the Proposers would encounter the Responders again down the line. This subconscious—and otherwise irrational— perspective coaxes us to defy the temptation to burn bridges when we are likely to cross them again in the future. In fact, neural imaging studies show that our natural penchant for cooperation is so ingrained that when we share, donate, and act fairly, the brain areas associated with reward are mobilized. We feel happy and satisfied as a result of our prosocial actions. Apparently, evolution has equipped us with a subconscious long-term perspective that often disagrees with the neoclassical economic assumption that we maximize utility in the here and now.

Now, why do so many people who play the Responder reject a 20% or even 30% offer? One answer is our evolved sense of fairness: we often instinctively gauge our gain based on relative payoffs against our rivals. Another answer lies again in reputation. When the Responder accepts a low offer—though seemingly better than nothing—his or her reputation suffers. Knowing his or her readiness to settle for less, others in the community may take advantage of the Responder in future transactions. As a result, in the long run, accepting low offers is costly. Rejecting low offers, on the contrary, is a worthwhile short-term sacrifice that serves to prop up a long-term public image. Clearly, rational choice theory has difficulties in handling such tradeoffs between short-term losses and long-term gains.

Evolution is a blind, amoral process, only caring about organisms that can leave behind the most copies of genes in their lifetimes, regardless of whether they behave selfishly or selflessly in the short term. That is, when the ultimate evolutionary prize is lifetime fitness, being either cooperative or competitive makes no difference, for they are the two sides of the same coin, serving the same end.

Humans have long lifespans and live in stable social groups. These conditions, as biologist Robert Trivers argues, favor cooperation, which often leads to better payoffs than what individuals can accomplish alone in the long run. That’s why prosocial behaviors have thrived in humans. Accordingly, rather than maximizing the utility in every one-shot deal as assumed in rational choice theory, being rational in evolution means behaving in a way that can lead to maximum lifetime payoff.

With the two conceptual modifications—relative and lifetime payoff—in rational choice theory, neoclassical economics can be made compatible with evolution and gain traction in the real world.

2016 April 12

Donating = Changing Economics. And Changing the World.

Evonomics is free, it’s a labor of love, and it's an expense. We spend hundreds of hours and lots of dollars each month creating, curating, and promoting content that drives the next evolution of economics. If you're like us — if you think there’s a key leverage point here for making the world a better place — please consider donating. We’ll use your donation to deliver even more game-changing content, and to spread the word about that content to influential thinkers far and wide.

 $3 / month
 $7 / month
 $10 / month
 $25 / month

You can also become a one-time patron with a single donation in any amount.

If you liked this article, you'll also like these other Evonomics articles...


We welcome you to take part in the next evolution of economics. Sign up now to be kept in the loop!

  • Derryl Hermanutz

    Excellent explanations! By adding just a little bit of real world complexity into the formerly simplistic understanding of rational behavior, the short term expense of cooperative behavior is shown to be clearly “rational”.

    I have long been uncomfortable with the related “delayed gratification” experiments: where a child’s ability to resist present benefit (eating a marshmallow) to gain even more future reward (getting two marshmallows) is interpreted as a predictor of life success. I see this more as a demonstration of the child’s “confidence in the future”. If the child lives in a highly competitive environment where competitors will grab any available marshmallows, then it is rational to eat one as soon as you get it; rather than saving it and expecting to get even more marshmallows in the future.

    So an alternate interpretation of delayed gratification is that if you live in a stable cooperative environment where your future feels secure because nobody is stealing your marshmallows, then your environment is preparing you for a successful life in a mutually rewarding cooperative world. But if you live in a highly competitive and insecure environment — eat, drink and be merry for tomorrow you die — then failing to seize the moment will produce present loss with no rational prospect of future gain.

    “Everybody” is better off in the cooperative world. And almost everybody is worse off in the competitive world: except the best competitors who are quick to grab and eat all the marshmallows.

    • Radical Philanthopy

      Your “delayed gratification” idea and mention of faith in the future is a good one because it also suggests the importance of environment. Research into the extremely tough childhoods of Romanian orphans has shown that they have biochemical differences in their brain, such as permanently higher levels of cortisol. It may be that children exposed to those types of stresses grow into adults physically hardwired to make decisions much more through short-term and relative payoff lenses as a survival strategy, which would be considered irrational to people from more typical childhoods.

      • L. Sun

        I think both of you are right about delayed gratification. On the one hand, it is stable social, economic, and political systems in today’s developed world that have made the future quite predictable. (E.g., a good education leads to statistically higher income.) Hence, our economic/political/social institutions favor those who opt for larger, later rewards (the future-oriented). On the other hand, when the future becomes more predictable, we are more confident in the future. That may be a reason why we plan far head and invest heavily on long-term projects (education, retirement, etc.). On the contrary, in a nation where the political/economic system is unstable (Syria, Sudan, Somalia, for instance), the future is unpredictable, and delayed gratification will become maladaptive. People tend to be more present-oriented, caring more about short-term rewards: they have to survive today first before thinking about living tomorrow. Therefore, it’s both our psychology and cultural environment that can make delayed gratification an either adaptive or maladaptive behavioral strategy.

  • Swami

    I am not sure if adding these two conceptual modifications fixes economics, but it does lead to a Rawlsian, or more appropriately, a Harsanyian form of social institutional choice.

    The societies which we would choose would be those which foster cooperation and competition aimed at improving absolute standards of living on an impartial basis. Today, people are choosing what can be summarized as the liberal state — free markets, strong government safety nets and scientific skepticism. The details are still up for debate, but broadly, this same model explains, Sweden, France, Japan, the US and increasingly China.

  • Rob Lewis

    Robert H. Frank covered the concept of relative (or “positional”) payoffs in his excellent book The Darwin Economy. Their existence completely overturns the “invisible hand” dogma, and (as Darwin understood) can produce individual incentives that are destructive for the group as a whole. The net effect is that we invest too much in positional goods (the classic “arms race”) and not enough in nonpositional ones. Government policies can correct this so we all end up better off.

  • This article is good as far as it goes, and Rob, Swami and Derryl all raise valid issues, which in my understanding can most powerfully be conceptualised in the following framework.

    Two major aspects to the framework – one aspect is the drivers of human choice, the other around what markets measure.

    On the choice aspect.
    All human beings are complex entities, capable of many different modes of interaction that are highly context dependent.
    One set of responses can be thought of in the broadest sense, as considering a spectrum of responses between short term selfishness and long term cooperation. In any given instant, how one responds on that spectrum seems to be a probability function, that has many sets of contributing factors including (but not limited to):
    how likely is it that there is enough of this thing for everyone?
    how essential is this thing to my survival (short and long term)?
    how likely is this thing to contribute to my enjoyment?
    if I cheat, how likely am I to get caught, and what are the consequences?
    what are the likely consequences of my actions now on my long term social success?
    how much time and energy is it likely to cost me to get some substitute for this thing?

    For most people, there is a large degree or randomness in even the seemingly simplest of choices. Few people maintain the sort of discipline that leads to the sort of consistency that results in thousands of successive choices without variation.

    Looking at what markets measure.
    Markets measure supply and demand. If supply exceeds demand, consistently, then price is zero.
    When most things were genuinely in short supply, that was a system of organisation and information transfer that had some real utility. As Hayek and others argued, the price signal led to successful cooperation – to a degree.
    A major problem with markets has always been that there exists a strong incentive to manipulate information about both supply and demand, to influence price (and thence profit).

    We are now in an age of automation, where we can duplicate and transmit information at a cost that is so low, that it is essentially free (some fraction of a cent for a month’s reading). The production of many physical goods is reaching the same sorts of levels of automation and production cost.

    This is where markets start to break down, and fail to deliver utility to the majority.
    Markets require scarcity to function.
    People require abundance to function.
    For a large and exponentially increasing set of goods and services we now have the technology to deliver abundance universally, but our system of markets cannot do that, as anything with universal abundance has zero price by definition (like air – arguably the most important thing to any human, yet of no market value).
    We could make all the essentials of life (air, water, food, shelter, education, healthcare, energy, communication, transportation, information) universally abundant (world wide), but no market system will ever do such a thing, in and of its own internal incentive structure.

    So here we are.
    We live in a market based world.
    We have the clear logical reality that markets have some major survival issues.

    We face a choice.

    To me, there is a way out, that does seem to work.
    Exponentially increasing technology does seem to offer the prospect of exponentially increasing abundance, and decreasing risk into the future.
    Classically uncertainty increases into the future, as unknowns overwhelm the known, leaving us with a discount rate on future benefits.
    It does seem possible now to conceptualise reversing that.
    It now seems likely that technology will be able to increase security into the far future, faster than uncertainty degrades it, which leads to a situation where cooperation is always a far higher pay-off than competition. Indefinite personal life extension, and increasing freedom, does seem to be a practical possibility.

    That is a clear possibility to me – just not so clear to many others as yet.

  • L. Sun

    Thank you all for your deep insights and a spirited discussion, especially from the economic perspective! My question for you is: are we able to use the concept of lifetime fitness for firms? While the relative payoff concept is straightforward, lifetime fitness for a firm is much more difficult to measure in practice. This is because a firm, unlike any biological organism, can live indefinitely. A large firm normally has a strong vitality and does not expect the end of its life any time soon. As such, it may have little concern on tradeoffs between short-term losses and long-term gains. As a result, large firms tend to be more selfish and less cooperative in short-term dealings (because they are less likely to perceive how short-term selfishness would affect their long-term economic prosperity). Can this be used to explain why Enron, BP, and VW did what they did? Or, am I in the wrong direction in my thinking?

    • Radical Philanthopy

      Firms are tricky because they look like independent entities but they include a multitude of individual actors working on different time frames and different incentives. One executive may be within a year of retirement and another just starting his career, but both could be involved in a firm’s decision-making process. Totally different incentives are in place for each actor in the firm and each actor’s view of his or her own role in the firm is a part of the picture.

      The firms you mentioned probably all had individual actors that were incentivized toward short-term rewards. In the Enron case, there was some evidence that executives knew the place was falling apart toward the end but sought to hide that knowledge so they could protect themselves at the expense of other members of the firm. (Executives quietly selling their own shares while encouraging lower-level employees to buy more.) That sort of behavior suggests you would have to look at Enron not as an individual firm but pool of different actors each doing what was rational in their own minds. They also had some group-think and internal cultural issues that may have trumped normal individual assessments of rational behavior.

      I think the research areas that would cover what you are considering would be organizational economics and organizational theory, which dig into how incentives, communications, structure, culture, and other organizational concepts tie into an organization’s behavior.

      • L. Sun

        Thank you for your suggestions! We’ve been working on how collective decisions are made, especially concerning the role of leadership in the process.

    • Onymous

      Good reply. The writer’s intuitions on this subject, that presumably led him to the insights shown in the article, show explicit reference to East Asian concepts of face and reciprocity, both of which are very important cultural norms in Asian societies. Interestingly, the institution in Chinese civilization most comparable to the Western corporation is not an economic institution, it is the political state. Traditional Chinese businesses, even of large scale (Wang An Computers in Massachusetts, Formosa Plastics in Taiwan, etc) are notorious for remaining personal fiefs of a strong founder and maybe his son in the second generation. Chinese just don’t trust impersonal processes (outside kinship based notions of filial subordination and reciprocity) that supposedly guarantee shareholder rights, as in Western corporations. The basic philosophy of the Chinese state is a cold Machiavellian realism about use of the ruler’s power over his subjects. (Remember, Machiavelli’s book was called “The Prince”.) Scholars call that Chinese school of statecraft the “School of Law”, which refers to regulations imposed impersonally and without mercy. Everyone’s security relies on the presence of an all powerful ultimate authority. The perennial hope is simply that the current despotism be an enlightened one. The Chinese state model has endured from the Warring States period, beginning around 400 BC, to the present, pausing only between strong centralized dynasties when regional forces jockeyed for supremacy. Whereas Marxism introduced the notion that the state is an epiphenomenal superstructure, called “upper storey architecture” in Chinese, standing on the fundamentals of economically defined class conflicts, the disastrous course of China’s Cultural Revolution discredited Marxist theory and left some people saying that, rather, in China the state itself is the “super-stable” fundamental reality that just doesn’t change. Contemporary China’s business practices play heavily on the allure of such notions as long term “friendship” and reciprocal granting of favors to maintain the relationship, as suggested in this article. However, as Bill Bishop has written in Sinocism newsletter, every business deal between a Chinese entity and a foreign entity must be approved by the Chinese government, and if the deal does not give the Chinese partner a clear advantage, it won’t be approved. Not to gain such advantage would be criticized as wasting Chinese people’s effort. The professed logic is that China is building now, and when China becomes strong, China will repay the favor, which means China will appropriate the foreign entity and beneficently bestow its superior cultural practices on the benighted foreigners. It’s amazing how many people fall for that, maybe as many as fall for Donald Trump’s personalized version of the same solipsistic narcissism. Presumably there’s an evolutionary basis for that kind of relationship, too.

  • David Quinn

    I can vouch for much of the sense of this. I used to work at a reasonably high managerial level for a large IT company. At the time many of the people in it had been there for a long time and many of the transactions were based on trust engendered by long association, for the reasons outlined here. The head of IT then got in a consultancy who tried to tell us that we should forget this clubbishness, as they put it, and zero in on stiletto sharp focus on our customers. Clearly these things are not incompatible but they said they were. When I left and became a consultant (contractor really) in another large IT department for a year I couldn’t get anybody to do anything for me, even mutually beneficial things. I formulated Quinn’s Law. People only do things for you if you are the boss or the boss’s factotum (in which case they will at least pretend to do it) or they have known you for 10 years (15 is better). The people around me would help each other. Current day management does not seem to understand this at all.

  • Onymous

    In the section titled “Rationality” is a Lifetime Deal, in the sixth paragraph, the words “the Proposer’s generality” should presumably be “the Proposer’s generosity”, right?

    • L. Sun

      Yes. Thanks for the correction and sorry for my oversight!

  • rorysutherland

    I think what you say is of immense importance in explaining much of the real world behaviour which is not explained by the standard economic model. Thank you for this.

    You quite rightly ask “How might a mentality tuned to such a long-term vision be favoured by natural selection?” You also highlight the importance of reputation. I suspect the two are highly connected.

    The effort of developing a mentality which is calibrated to consider long-term pay-offs is only worthwhile if it is possible reliably to signal to others that you have such an inclination. If you operate to the rules of a long temporal frame, but are treated by others as though you are merely interested in quick wins, all your delayed gratification is wasted.

    Therefore a large part of anyone’s reputation must be derived from reliable (ie “costly”) forms of signalling which prove that they are conducting their life according to the rules of the long-term game, not the very different rules of short-term self-interest. Behaviours which are reputation-enhancing would therefore most likely be those behaviours which are immediately costly but which pay back only over a longer time-frame.

    In human terms, buying an engagement ring is evidence of long-term intent and faith in an enduring relationship. (If you were planning a one-night stand, you would be reluctant to invest two-months’ salary in a preliminary present.) Going to university is a reliable way to signal to an employer that you are prepared to take a hit now in exchange for gains later. Advertising your product in costly media is a reliable way of signalling that you believe the product will not be a flash in the pan, since it would not pay to advertise products which would not be enduringly popular. Refreshing the decor of your hotel proves that you have a more far-reaching commitment to your business than someone with a shabby hotel milking its convenient location for short-term gain from non-repeat customers. People and businesses which do not invest in such costly signalling devices unconsciously suggest they do not care about the future much, and so are inherently less trustworthy. (I once turned up at a restaurant to see a sign in the window announcing that it was shortly to close – instinctively my heart sank. Why should they make me a good meal if there were no prospect of repeat business?) Good manners are generally effortful – and so are a reliable means of proving that you care what other people think of you. Again, someone with a short time-horizon would not bother opening doors for people, writing thank-you letters and so forth.

    One interesting product of standard economic theory is the “shareholder value movement” which tends to shift businesses into an obsession with short-term returns. This may be hugely damaging to the creation of reputational capital, and to the premium customers are willing to pay to deal with someone who visibly prefers to make ten fair and moderately profitable repeat sales to one customer to ripping off ten customers once.