Book Reviews 797 economics profession embraced morality as a In biology, the theory of evolution indeed creates worthy area of research? Are publications on sense where there was none before. morality and economic outcomes getting pub- To say the same about economics and evolu- lished in A-level journals? To answer these ques- tion would of course be too much. Many things tions, I conducted a couple of searches on the in economics make sense and will keep on making distinct words “morality,” “culture,” or “trust.” sense with or without evolution. But since eco- First, I undertook a narrow search using nomics is all about human behavior, and humans JSTOR. I searched the titles and abstracts of have become what they are through mutation and articles between 1970–2012 appearing in the selection, one could imagine that, if there should American Economic Review, Quarterly Journal be a fundamental source for hypotheses concern- of Economics, and Journal of Political Economy. ing how we behave, the only serious candidate is In all, I found 2 hits for “morality,” 17 for “cul- evolution. Of course there is also rationality, but ture,” and 26 for “trust.” My second, broader that is only an intermediate source for hypotheses search relied on all economics and finance jour- since a natural follow-up question would be how nals in ScienceDirect since 1970. The numbers and why evolution would have selected for ratio- were revealing. There were 59 captures on nality, or why it would have selected the (bounded) “morality,” 795 on “culture,” and 657 on “trust.” rationality that we have. It was therefore only a There is a decided structural break around 2008 matter of time that an interest in evolution would (presumably owing to the financial crisis) when arise among economists, just as it would—and the frequency of hits nearly doubles. The Moral has—in psychology, sociology, and anthropology. Foundation of Economic Behavior, I hope, will Samuel Bowles and Herbert Gintis were among help hasten this change in the profession. the first from economics, and their A Cooperative Species: Human Reciprocity and Its Evolution References reflects around ten years of work exploring evolu- Wilson, James Q. 1993. The Moral Sense. New York: tionary explanations of how we got to be the coop- Simon and Schuster, Free Press. erative, reciprocal animal­ that we are. Janice Boucher Breuer The quest for an evolutionary explanation of University of South Carolina human cooperation of course did not start at the turn of the century; it naturally came right along with Darwin’s great insight. It is only relatively D Microeconomics recently that economists became interested in evolution too. In the 1990s, there was a wave of A Cooperative Species: Human Reciprocity and research in evolutionary game theory, which was Its Evolution. By Samuel Bowles and Herbert mainly focused on disequilibrium dynamics and Gintis. Princeton and Oxford: Princeton how learning could make behavior converge to University Press, 2011. Pp. xii, 262. $35.00. equilibrium play. One of the nice spillovers of ISBN 978–0–691–15125–0. that first wave is that it introduced us to tools JEL 2011–1324 such as the replicator dynamics and to equi- librium refinements such as the evolutionarily Introduction stable strategy, which are also useful for evolu- tionary theory that is not limited to the within In 1973, geneticist and evolutionary biologist generations timescale one typically associates Theodosius Dobzhansky wrote an essay in which with learning models. Thereby it prepared us for he argued against what we now would call intel- a second wave of evolutionary explanations, this ligent design. To underscore his position about time not only of how we update our behavior, how theory and empirical evidence relate, he given what we want and what we value, but also gave the essay the famous title Nothing in Biology how we came to want and value those things in Makes Sense Except in the Light of Evolution. the first place. That of course makes for a fantastically power- The book by Bowles and Gintis reflects their con- ful quote and it has been used a million times. tributions to this second, more interdisciplinary 798 Journal of Economic Literature, Vol. L (September 2012) wave. It draws on journal articles of theirs, departures from random group formation. together and apart, with and without other coau- Because that implies that individuals within thors. These papers are reorganized, rewritten, groups are (somewhat) related, it is natural to and enriched with chapters that provide a back- wonder whether group selection and kin selec- ground. Although the seams and patches are still tion are one and the same thing, or if perhaps somewhat visible, it is very welcome as a guide to, one is subsumed under the other. The answer and overview of, their work. to that question is somewhat involved, but it is safe to say that group selection should come with Keen to Cooperate, Eager to Punish some degree of relatedness, and one of the inter- esting follow up questions is of course whether In the description of the behavior that is or not the conditions for normal, run-of-the-mill to be explained, Bowles and Gintis contrast group selection to explain our levels of coop- the empirical evidence with the tendency, eration are satisfied in humans. Therefore they observed in economists and Americans, to try introduce the reader to the ​FST​ ​ , which is a mea- to explain all cooperative behavior as a result sure of relatedness. of self-interest, rightly understood. They con- They then go on to describe how reciproc- vincingly argue that such an explanation would ity features in the repeated games literature in not do justice to the keen eye for cooperation economics. The gist of their discussion is that that humans have, their tendency toward fair- we have been very creative in finding equilibria ness, and their taste for retribution. Of course, in repeated games, squeezing the shadow of the we have a keen eye for our own interests too, future dry in a variety of settings, but that the but sometimes we are simply genuinely happy existence of equilibria is not the same as them to help out and intrinsically eager to punish being likely to be established. The more players free-riders. The obvious next question is then: there are, and the less information players have why? Before going to their own contributions, about past actions, the harder it gets for us to find they first describe the existing, classical expla- equilibria, let alone equilibria that we can expect nations from sociobiology. There is kin selec- are also easily found by a process of mutation and tion theory, which implies that selfish genes can selection. They argue that there is a role to be make altruistic people because interactions may played for norms here. take place between related individuals. If indi- viduals are related, then genes that make their Relatedness and Warfare bearers confer a benefit to the other, at a cost to themselves, can be selected for because of the In the next chapter, they venture the farthest fact that these benefits are conferred, not just from their home ground. They give estimates to an average individual from the population, of measures of relatedness for a number of but to an individual that has an increased prob- hunter–gatherer populations, and conclude from ability of being a carrier of the very same gene. those that, in the absence of other ingredients, Of course, in order for this to work, the donor “ancestral conditions were not favorable for the and the recipient have to be sufficiently related, ­evolution of group-level cooperation by means given costs and benefits. of reciprocal altruism” (101). The link between

A second classic is group selection. The short- estimated ​FST​ ​’s and reciprocal altruism came a est summary of this process is from Wilson and bit as a surprise to me. For given costs and ben- Wilson (2007), who state that “Selfishness beats efits in a standard simple model, it is clear what altruism within groups. Altruistic groups beat a sufficiently high relatedness for unconditional selfish groups. Everything else is commentary” cooperation to evolve would have been; that is, (345). The balance between those two oppos- given by a famous criterion known as Hamilton’s ing forces—within group selection and between rule. Where the threshold for conditional coop- group selection—determines whether altruism eration lies is not clear. The authors claim that the ​ or selfishness wins. The conditions for group FST​ ​’s are not particularly large, and repetitions not selection to work in favor of altruism involve impressively frequent, but from the book it is not Book Reviews 799

clear at which estimates of F​ ​ST​’s and continuation group selection pressure against cooperators. probabilities the conclusion would have been that Norms are also assumed to be a group property, this is, in fact, sufficient for conditional coopera- which is retained if groups fission. Because of tion to readily evolve. But even though this point their effect on within group selection, correlation remains somewhat less sharp, it is only the upbeat between the level of cooperation and the level of to one of the main themes in the book, which the norm builds up over time, and because the loses none of its appeal. The key ingredient that more cooperative groups beat the less coopera- Bowles and Gintis suggest has played a major role tive ones, norms can spread. Norms and behav- in the evolution of human cooperation is, para- ior therefore are coevolving; they play leapfrog, doxically, intergroup warfare. Before they go on the norm making cooperation less costly, and to models of intergroup conflict, they provide us the increased cooperation being what makes the with archaeological and ethnographic evidence of norm spread by winning wars for the groups that mortality due to warfare. These empirical studies have them. suggest that violent intergroup interactions were After the extension with norms, Bowles and a fact of life for our ancestors, and Bowles and Gintis go back to the model without norms, but Gintis point out that this would drastically alter take that one step further by endogenizing inter- the selective forces concerning cooperation. group conflict. Individuals are endowed with two The authors deserve a lot of credit for that characteristics: their readiness to engage in con- contribution. They are the most prominent radi- flict with other groups—which they call parochi- cal economists from the 1970s—which I guess alism1—and their level of altruism. Again, these makes them the most prominent radical econo- two traits coevolve; if parochialism in the overall mists of all times—and as such they are expo- population is high, then conflicts are regularly nents of the generation that was all about love, occurring and, if they are, within group coopera- peace, and understanding. The message that tion is paying off. Unlike the coevolution of norms violent conflict shaped human nature will feel and cooperation, here there are two distinct clus- uncomfortable to most of us, and I guess it will sit ters of states that are relatively stable; one with even less well with their demographic. But their low parochialism and low altruism, and one with inconvenient truth is pretty convincing. If compe- high parochialism and high altruism. The popu- tition between groups is by direct confrontation, lation sometimes makes transitions between the and not in more indirect ways, such as through two clusters, and then typically remains within a pressure on scarce resources, then that is a real cluster for a while. game changer. Conflict between groups aligns the interests of its members. A war is won or lost A Dish Best Served Collectively collectively, so it becomes a matter of united we stand, divided we fall. What would have been a The most interesting material in this book is all costly act of altruism in the absence of war might a little over halfway. The chapters on warfare (7 become a less costly one, or even an act that and 8) tie for being the best part with the chapter serves the actor’s own interests too. In a way, one on coordinated punishment (chapter 9). In ear- could say that, somewhat contrary to the initial lier chapters, the authors have argued that Folk message, it brings cooperation a bit closer to self- Theorem equilibria, and especially the intricate interest, rightly understood again (see Van Veelen punishing strategies, are not necessarily particu- and Hopfensitz 2007). larly realistic, especially not if there are more Apart from the direct effect on cooperation than two players involved (they refer to them as of this particular form of competition between evolutionarily irrelevant equilibria). In chapter 9, groups, Bowles and Gintis also consider models with intergroup conflict in which they allow for 1 This definition of parochialism is from Choi and coevolution of institutions, or norms, and cooper- Bowles (2007). It is not the same as in Bernhard, Fischbacher, and Fehr (2006), where a parochialism is ative or altruistic behavior. Norms within a group someone who is nice to in-group members and not nice to can have an effect on within group selection, and out-group members. See also García and van den Bergh the norms they consider typically reduce within (2011). 800 Journal of Economic Literature, Vol. L (September 2012) they suggest evolutionarily more relevant ones. would like to zoom in on the first few formulas of They do that by adding the possibility to dish the book. They are quite representative, because out targeted punishment to individuals. This way lots of the material later in the book is related to punishment by withholding cooperation is no lon- those. In order to do that, I have to introduce the ger the only possibility. Withholding cooperation Price equation. This equation plays a peculiar role normally hurts everyone, including those that did in theoretical biology. It is named after George R. cooperate, and coordinating on hurting one indi- Price and while a part of the field claims it cap- vidual is difficult. What happens in their equilib- tures the essence of selection in full generality, ria is that, before a public goods game is played, others—including me—think the Price equation players check each other out. In this phase, they is more like the emperor’s new clothes than the can signal that they are, in principle, ready to holy grail. The majority, however, cannot really punish those that do not cooperate. The strate- choose between the two, and regularly gives it the gies they consider are (1) nonpunishers and (2) benefit of the doubt. punishers. The nonpunishers do not punish, but Perhaps the Price equation can best be intro- they are sensitive to punishment; if they expect duced in the words of the famous evolution- sufficient punishment to make cooperation a ary biologist and inventor of evolutionary game more attractive option, they will cooperate. The theory John Maynard Smith. In an interview punishers do punish if they find themselves in a in which he comments on George Price and group with at least a certain minimum number of their work together, Maynard Smith says: “I am other punishers. And they cooperate under the not going to tell you what Price’s Theorem2 is, same conditions that nonpunishers do. because I don’t actually understand it.” He then This setup allows for two stable equilibria—a goes on to explain that he thinks it offers an alter- pure one of nonpunishers only, and a mixed one. native route to understanding evolution, but one What I like about the model is that it brings us that he never takes. His reaction to the Price one step closer to the rich communication that ­equation is a very natural one, and most people is typically there in the formation of normative will remember something similar from their first judgments. Lots of our daily talk is about behavior few encounters; typically people are not really we approve or disapprove of in others, and agree- sure they understand it, but assume that others ing about right and wrong makes morality much understand something about evolution with it in more of a rhetoric business than we usually model a way they do not themselves. it to be. The self-serving bias and our extensive My claim is that the Price equation is an identity rhetoric in our moral judgments attest to the that does not help derive results, but that can cre- fact that, as psychologist Jonathan Haidt (2012) ate a premature sensation of understanding at a claims, we evolved to be lawyers. The stable mix- point where the work is not yet done (Van Veelen tures are also a nice feature. And even though 2005). This would not go undetected if the format the strategy space in this model is still not super for theory would be that claims have to be pre- rich, the model is a few steps in the right direc- sented as theorems with proofs. Unfortunately tion from models with strategy sets that were not this is not the norm in theoretical biology, and particularly robust to enlargements of those strat- the literature contains many derivations of results egy sets. Plus, our strong reciprocity (costly pun- with the Price equation. On closer inspection, ishment of defectors in public goods games) is in these results sometimes turn out to be correct— need of an explanation, and this could be one. although the derivation with the Price equation is no proof—but more often they turn out to be true The Price of Using the Equation only under (additional) assumptions. On some

There are also lesser aspects of the book. They 2 lie not so much in the intuition behind the ideas, It is interesting and telling already that John Maynard Smith and others sometimes use the word theorem instead but more in the technical presentation. As a of equation or identity, even though there is no such thing prime example of how technical imperfections as Price’s Theorem in the literature. The only thing that is hinder smooth comprehension of the material, I there is an identity; see also Van Veelen et al. (2012). Book Reviews 801 occasions, they are just wrong, or ill-defined, altruist, conditional on the first being one, and the and therefore not even wrong (Van Veelen et al. probability that the second individual is an altruist, 2012). conditional on the first not being one. The equations from the book by Bowles and (4.9) ​F​ ​ (A A) (A N) Gintis that I want to zoom in on are (4.7) to (4.9). ST = 핇 | − 핇 | The first is arrived at after a few manipulations It is intuitive that this should matter; if altruists of the Price equation, and is meant to capture are more likely to interact with altruists than non- the balance between within group selection and altruists are, that helps. between group selection. One remarkable aspect of those two definitions

is that the range of the ​F​ST​’s that they allow for (4.7) p (b c)var( ​p​j​) c ​var___​( ​pij​​) is different. From the definitions it immediately Δ = − − follows that the F​​ST​ from (4.8) varies between 0 and 1, while the F​ ​ ​ from (4.9) can take values The idea behind this formula is simple. The bene- ST between –1 and 1. That is peculiar, and makes fitsb minus the costs c of the act of altruism is the one want to check if they really are the same for efficiency gain for a group of an altruistic act. The a simple example. So let’s assume that groups are larger it is, the better for altruists. Then var( ​p​​) is j of size 2. One out of four groups consists of two the between group variance, so if this is large, nonaltruists, one out four consists of two altruists, and groups are very different in composition, and two out of four consist of one altruist and one this intensifies the selection between groups, nonaltruist. Using the definitions in the book,3 which is also good news for altruists. On the that makes var(​p​​) 1/8 and ​var___​(​p​​) 1/8. With other hand, if costs of cooperation c are large, j = ij = (4.8), that gives an ​F​ ​ of 1/2. According to (4.9), that is bad for altruism. Moreover, if groups ST however, ​F​ ​ 0. That is a massive difference. ___ ST = are heterogeneous, and ​var ​( ​pij​ ​) is large, within group selection is strong, which is again bad news The population example was chosen to closely for the evolution of altruism. resemble what random matching would result We will return to equation (4.7) in a bit, but in in large populations. We know that random first I would like to give the other two equations, matching should match a relatedness of 0, which which are presented as two equivalent definitions makes one tend to go for the second definition. of Wright’s inbreeding coefficient. The first one however is linked to equation (4.7), so it is worthwhile to also explore how this ran- dom matching example would work out in that var(​p​j​) (4.8) ​F​ ​ ​ equation. With var(​p​​) 1/8 and ​___var​(​p​​) 1/8, ST = __​var___​(​p​​) var(​p​​) j = ij = ij j equation (4.7) becomes p b/8 c/4, which + Δ = − would imply that with random matching in This one uses variances that also feature in equa- large populations, the frequency of altruists will tion (4.7). The next one uses conditional probabili- increase if benefits are at least twice the costs. ties. These conditional probabilities are thought This contradicts all benchmark results from evo- of as pertaining to a simple chance experiment: lutionary game theory and evolutionary biology, first an individual is drawn from the population as which state that with random matching behav- ior never spreads if c 0, or, in other words, if a whole, with every individual equally likely to be > drawn, and then a second individual is drawn from the same group—without replacement—where 3 For simplicity, we assume that groups all have the all other group members are equally likely to be same size n, and that there are m groups. The definitions drawn. The probability with which the second in the book are not restricted to groups of equal size, but here that is not important. ​p​​ is 1 if member i of group j is individual is an altruist may depend on whether or ij an altruist, and 0 if it is not. ​p​j​ is the average value of ​pij​​ in 1 n not the first individual is one; if groups are formed group j; ​p​j​ _n ​ ​ ​ i 1​ ​pij​ ​ , p is the average value of ​p​ij​ in the = Σ = 1 m assortatively, it will. The alternative definition of entire population; p _ m ​ ​ ​ j 1​ ​pj​ ​ . Between group variance = 1 Σ = 2 the F​ ​ST​ reflects that: it is the difference between is defined as var(​pj​​) _m ​ ​ ​ j​ (pj​​ p​)​​ and average within = Σ − 1 2 group variance is defined as ​___var​(​p​​) ​ ​ ​ (p​ ​​ ​p​​)​​. the probability that the second individual is an ij = _nm Σ​ ij ij − j 802 Journal of Economic Literature, Vol. L (September 2012) it is costly and therefore truly altruistic (see for direction of selection is a big controversy in evo- example Van Veelen 2009). lutionary biology. But whichever side is right, The first set of formulas therefore puts the they both will agree that inclusive fitness is not a inquisitive reader in a bit of a difficult situation. label for a model with assortment, but a quantity If one decides not to dig deeper, equation (4.7) that may or may not be a good indication for the reflects a perfectly OK intuition for what the direction of selection. things are that matter for group selection. But if The authors also claim a few times that, in their one does act on curiosity, one finds out that equa- classical paper, Axelrod and Hamilton (1981) tion (4.7) is actually not correct, and also the two “show that a modest amount of kin-based assort- supposedly equivalent definitions of the ​FST​ ​ turn ment encouraged a reciprocal altruist behavior to out not to be the same. Both the Price equation emerge when rare and proliferate” (114). That is and the ​FST​ ​ regularly return later in the book, and not true; their analysis is based on random match- for instance at the point where conclusions are ing, and they only verbally suggest that assort- drawn from estimates of the ​F​ST​ it seems infea- ment could give tit-for-tat the right kick-start. sible for the reader to go over those estimates too, Another remarkable detail is that the authors trying to decide whether or not the right definition use annual mortality rates to determine what the was used, and if not, how that affects the conclu- relevant continuation probability has been in the sions that are drawn. The casual notation also does typical repeated games in our evolutionarily rele- not make it particularly easy to dig deeper, and I vant past. Thereby they seem to implicitly assume think the natural response of many readers will be that the relevant games are played once a year. to give up on these equations at some point. That may be the case for some situations where there are gains from cooperation, but not for all. Other Technical Issues Opportunities for collective hunting may present itself more often than once a year, while help to There are a few more examples of technical rebuild a home may be required only every once imperfections. They typically are caused by a in a blue moon.4 somewhat loose treatment of the classics in evo- lutionary biology. Readers that are already famil- Conclusion iar with those will be able to make up for those lapses of accuracy, or maybe even skip over them A fair share of the work that needs to be done altogether, but for readers that are not immersed in science will always have to be focused on in the evolutionary literature, extra care and pre- ­formulas and theorems. Yet I guess one can safely cision is always welcome. argue that real progress is made not so much by The authors are, for instance, a bit unorthodox detailed tuning of the math as it is by introducing in their use of the term inclusive fitness. They bold new ideas. And Bowles and Gintis are clearly say that “Following Hamilton we apply the term not short of ideas. The attention they draw to the inclusive fitness to all models in which positive role of conflict and coordinated punishment in assortment allows the proliferation of an altru- the evolution of our cooperative and reciprocal istic trait” (48). They apparently use inclusive species makes the book very much worth read- fitness as a name for a type of model. That mis- ing. Their focus on the evolution of human nature represents what inclusive fitness is. Inclusive also paints a much richer picture of our behav- fitness is a quantity that can (or sometimes can- ior than traditional economics tends to do. They not) be computed for a given behavior in a given could however have done their readers an addi- model. The idea behind it is that the prediction tional favor if they had treated the presentation of for such a model follows the value of that quan- their ideas with a little more care. Nothing about tity; the (altruistic) behavior is selected for if human nature makes sense except in the light of inclusive fitness is larger than 0 and is selected evolution, that is for sure, but it never hurts to against if inclusive fitness is smaller than 0. bring some good math along. Whether or not there is a meaningful defini- tion of inclusive fitness that always matches the 4 Dave Rand pointed this out to me. Book Reviews 803

References book, and makes the central point that happiness Axelrod, Robert, and William D. Hamilton. 1981. data can add to the understanding of individual “The Evolution of Cooperation.” Science 211 well-being provided by the analysis of income and (4489): 1390–96. revealed preferences. Bernhard, Helen, Urs Fischbacher, and Ernst Fehr. Chapter 2 attempts to set out what we mean by 2006. “Parochial Altruism in Humans.” Nature 442 (7105): 912–15. happiness. A broad distinction is made between Choi, Jung-Kyoo, and Samuel Bowles. 2007. “The a fulfilling life (with agency or capabilities to be Coevolution of Parochial Altruism and War.” Sci- able to change one’s life) and contentment. These ence 318 (5850): 636–40. are called the Aristotelian and Benthamite views Dobzhansky, Theodosius. 1973. “Nothing in Biology of happiness, although the academic literature Makes Sense Except in the Light of Evolution.” American Biology Teacher 35 (3): 125–29. often refers to Eudaimonic and Hedonic well- García, Julián, and Jeroen C. J. M. van den Bergh. being, respectively. Most surveys ask some kind 2011. “Evolution of Parochial Altruism by Multi- of hedonic question, on happiness, satisfaction, level Selection.” Evolution and Human Behavior or Cantril’s “best possible life” (ladder) question. 32 (4): 277–87. Haidt, Jonathan. 2012. The Righteous Mind: Why An open question in the empirical literature is the Good People Are Divided by Politics and Religion. degree to which meaning and capability, on the New York and Toronto: Random House, Pantheon one hand, and happiness, on the other, are cor- Books. related. J. S. Mill’s happy pig suggests that these van Veelen, Matthijs. 2005. “On the Use of the Price are not synonyms; but in practice they could be Equation.” Journal of Theoretical Biology 237 (4): 412–26. quite closely matched. It is however possible that van Veelen, Matthijs, Julián García, Maurice W. Sabe- acquiring agency may lead to lower happiness, at lis, and Martijn Egas. 2012. “Group Selection and least in the short run. As such, potentially valuable Inclusive Fitness Are Not Equivalent; The Price long-run policies may be initially disruptive, and Equation vs. Models and Statistics.” Journal of Theoretical Biology 299: 64–80. individuals certainly dislike uncertainty. Equally, van Veelen, Matthijs, and Astrid Hopfensitz. 2007. too much happiness may reduce any incentive “In Love and War: Altruism, Norm Formation, and for action. Overall, there are a number of intrigu- Two Different Types of Group Selection.” Journal ing points here, although the development is not of Theoretical Biology 249 (4): 667–80. always particularly easy to follow. Wilson, David Sloan and Edward O. Wilson. 2007. “Rethinking the Theoretical Foundation of Socio- Carol Graham is on surer ground in chapter 3, biology.” Quarterly Review of Biology 82 (4): on Happiness around the World, which draws 327–48. on some parts of her recent Oxford University Matthijs van Veelen Press book of the same name. Much research University of Amsterdam has concerned developed countries; but most people don’t live in developed countries. Cross- country empirical analysis reveals that some of I Health, Education, and Welfare the happiness correlates are very similar across countries . . . but not all of them. In particular, The Pursuit of Happiness: An Economy of Well- the effect of labor-force status suggests that there Being. By Carol Graham. Brookings Focus may be more choice in this respect in developed Books. Washington, D.C.: Brookings Institution relative to developing countries. Overall, this is a Press, 2011. Pp. vii, 164. $24.95. ISBN 978–0– useful potted history of what happens when we 8157–2127–7. JEL 2011–1462 ask identical well-being questions to individuals Brookings’ Focus book series aims to present in very disparate countries. policy issues of interest to a broad audience. This Chapter 4 deals with adaptation and social short book by Carol Graham contributes to this comparisons. If either pertain, then it is not series, and presents a number of topics in the enough to know how much X an individual has burgeoning economics of well-being. to evaluate their well-being; we also need to The book is divided into five chapters, each of know how much X others in the reference group which has extensive endnotes. The first previews have (social comparisons), or how much X the a number of the issues that will appear later in the individual used to have in the past (adaptation).