W06-HUANG-WITH FIGURE 04/07/01 6:36 PM HERD BEHAVIOR IN DESIGNER GENES ∗ Peter H. Huang The ability of individuals to choose their children’s genes has increased over time and may ultimately culminate in a world involving free market reprogenetic technologies. Reprogenetic technologies combine advances in reproductive biology and ge- netics to provide humans increased control over their children’s genes. This Article offers economic perspectives that are help- ful in understanding the possibly unexpected ethical, legal, and social issues at stake in using reprogenetic technologies for trait enhancement selection. The Appendix analyzes two competitive games that might arise in such a biotechnological society. Spe- cifically, the Article focuses on herd behavior, caused by either a popularity contest or positional competition, in the choice of genetic traits. The analytical game-theoretic models in the Ap- pendix can have several equilibrium outcomes in terms of indi- vidual reprogenetic technological choices and corresponding be- liefs about such choices by others. This multiplicity of potential social outcomes suggests that a society can attain effi- ciency if the state or some private organization transforms in- dividual parents’ beliefs over the choices of other parents re- garding their children’s traits and, thus, coordinates parental reprogenetic decisions by selecting, as focal, certain beliefs over parents’ reprogenetic decisions. TABLE OF CONTENTS INTRODUCTION..........................................................................................640 I. SOURCES OF HERD BEHAVIOR FROM INDIVIDUAL REPROGENETIC CHOICE.............................................645 II. ROLES FOR LAW AND PUBLIC POLICY.............................................653 III. OTHER CONCERNS WITH DESIGNER GENES...................................657 ∗ Assistant Professor of Law, University of Pennsylvania. A.B., 1976, Princeton University; Ph.D., 1984, Harvard University; J.D., 1997, Stanford University. Many thanks to Scott Altman, Anita L. Allen-Castellitto, Rachel Croson, Oliver Goodenough, Hank Greely, Mark Hall, Pamela Harris, Jason Johnston, Leo Katz, Russell Korobkin, Eric Posner, Amanda Spitzer, Matt Spit- zer, Eric Talley, and the audience members of the summer 1998 Wharton Ap- plied Economics Lunch and the Gruter Institute for Law and Behavioral Re- search 1998 Conference on The New Genetics and Reproduction: The Legal Response for their helpful comments and suggestions. 639 W06-HUANG-WITH FIGURE 04/07/01 6:36 PM 640 WAKE FOREST LAW REVIEW [Vol. 34 CONCLUSION.............................................................................................662 APPENDIX: TWO HERD MODELS OF REPROGENETIC SELECTION..............665 Never forget who you are / Little star / Never forget how to dream / Butterfly / God gave a present to me / Made of flesh and bones / My life / My soul / You make my spirit whole / Never forget who you are / Little star / Shining brighter than all the stars in the sky / Never forget how to dream / Butterfly / Never forget where you come from / From love / You are a treasure to me / You are my star / You breathe / New life / Into my broken heart1 INTRODUCTION Reprogenetic technology combines advances in reproductive bi- ology with genetics to provide individuals increased control over their children’s genetic traits.2 As an editorial from The Economist argued, “The proper goal is to allow people as much choice as possi- ble about what they do. To this end, making genes instruments of such freedom, rather than limits upon it, is a great step forward. . . . [B]iology will be best when it is a matter of choice.”3 As with other forms of technology, however, a society can limit or even ban the use of designer genetics if there is reason to fear problems from its us- age. Such techniques as human cloning, embryo selection, and ge- netic engineering raise many subtle and complex ethical, legal, and social issues (“ELSI”). This Article focuses on one of these issues, namely, the possible types of herd behavior that can occur in a free market world in which parents can delete or insert genes of their choice into their children. One should notice that this Article is not entitled “designer traits,” “designer babies,” “designer children,” or “designer adults.”4 These titles were not chosen because the majority of traits, and cer- tainly those related to complex behavior, are polygenic5 and multi- factorial.6 Genes are often described as “blueprints”7 because they 1. MADONNA, Little Star, on RAY OF LIGHT (Warner Bros. Records 1998). 2. See LEE M. SILVER, REMAKING EDEN: HOW GENETIC ENGINEERING AND CLONING WILL TRANSFORM THE AMERICAN FAMILY 9 (1998) (defining reprogenet- ics). 3. Changing Your Genes, ECONOMIST, April 25, 1992 at 11, 12. 4. Michael D. Lemonick, Designer Babies, TIME, January 11, 1999, at 64- 65 (discussing and reporting on surveys about designer genetics). 5. See MAXWELL J. MEHLMAN & JEFFREY R. BOTKIN, ACCESS TO THE GENOME: THE CHALLENGE TO EQUALITY 17-18 (1998) (defining polygenic traits as those for which more than a single gene is required for the trait’s transmission). 6. See id. (defining multifactorial traits as those produced by both genetic and environmental factors). 7. Claudia Wallis, Can Prayer, Faith and Spirituality Really Improve Your W06-HUANG-WITH FIGURE 04/07/01 6:36 PM 1999] DESIGNER GENES 641 are said to contain “instructions for making various proteins.”8 The metaphor of a blueprint, however, implies a misleading level of pre- cision and rigidity.9 A better and more appropriate analogy is that genes are like cake recipes, which are simply a set of instructions that should, if followed properly, produce a cake or cake-like item.10 Of course, as most people who have baked cakes or attended a cooking school will admit, there is much variation in the cakes cre- ated from an identical recipe. In the language of genetics, the in- structions are the genotype, while the results are the phenotype. Genes do not determine phenotype.11 Identical, or monozygotic, twins have the same genetic information, but develop quite differ- ently when exposed to different environments.12 Another useful analogy comes from playing cards, where players receive genetic hands, “but only in very rare cases does this automatically define a player as a winner or a loser. Rather, any outcome depends on how the cards are played, which in turn depends on how that play inter- acts with the playing of cards from other hands beyond the original player’s control.”13 The possibility of parents being able to alter their children’s ge- netic make-up poses difficult and controversial ELSI and philo- sophical issues. For example, prenatal screening through amniocen- tesis may indicate the presence of a fetal genetic disease. A couple’s subsequent decision to abort the fetus raises the issue of what con- stitutes a true deformity, as opposed to a mere difference.14 Other Physical Health? A Growing and Surprising Body of Scientific Evidence Says They Can, TIME, June 24, 1996, at 58, 61 (discussing the possibility that hu- mans possess a genetic blueprint for believing in religion). 8. ERIC S. GRACE, BIOTECHNOLOGY UNZIPPED: PROMISES AND REALITIES 20 (1997). 9. See TIMOTHY H. GOLDSMITH, THE BIOLOGICAL ROOTS OF HUMAN NATURE: FORGING LINKS BETWEEN EVOLUTION AND BEHAVIOR 70-72 (1991) (describing the myth of biological determinism); Henk Jochemsen, Reducing People to Genetics, in GENETICS ETHICS: DO THE ENDS JUSTIFY THE GENES? 75, 77 (John F. Kilner et al. eds., 1997) (describing how the central dogma of molecular biology implies a genetically deterministic view of humans); Michael Ruse, Knowledge in Human Genetics: Some Epistemological Questions, in GENES AND HUMAN SELF- KNOWLEDGE 34, 38-42 (Robert F. Weir et al. eds., 1994) (discussing the dangers of methodological reductionism with the Human Genome Project). 10. See MARTIN BROOKES, GET A GRIP ON GENETICS 178-83 (1998) (providing the metaphor of cake recipes for genes). 11. See RICHARD LEWONTIN, HUMAN DIVERSITY 19 (1982) (discussing the common error that genes determine phenotypes). 12. See ROBERT PLOMIN ET AL., BEHAVIORAL GENETICS: A PRIMER 309-60 (2d ed. 1990) (describing twin studies and adoption designs); NANCY L. SEGAL, ENTWINED LIVES: TWINS AND WHAT THEY TELL US ABOUT HUMAN BEHAVIOR 20-35 (1999) (describing research on identical twins). 13. Charlie Davison, Predictive Genetics: The Cultural Implications of Sup- plying Probable Futures, in THE TROUBLED HELIX: SOCIAL AND PSYCHOLOGICAL IMPLICATIONS OF THE NEW GENETICS 317, 319 (Theresa Marteau & Martin Ri- chards eds., 1996). 14. See Evelyn Fox Keller, Nature, Nurture, and the Human Genome Proj- W06-HUANG-WITH FIGURE 04/07/01 6:36 PM 642 WAKE FOREST LAW REVIEW [Vol. 34 examples of reprogenetics include a dwarf couple who wants their child to be a dwarf15 or a deaf couple who wants their child to be deaf.16 In addition, there are constitutional and equity (in particu- lar, affordability) issues raised by the prospect of a free market sys- tem of reprogenetics.17 This Article does not focus on these inter- esting and important issues, nor does it deal with eliminating life- threatening or other so-called diseases. Instead, this Article focuses on genetic selection of what have been termed enhancements, such as height, weight, hair color, skin color, intellectual ability, or possi- bly pre-dispositions toward certain types of behavior. Although “pa- rental eugenics,” or, more descriptively, “homemade eugenics,”18 are not currently feasible, they may be in the near future. One choice currently available to parents is the sex of their children. As Schel- ling noted, gender may serve as an effective proxy for other charac- teristics over which parents have preferences.19 It may seem that the ability to select the genetic traits of one’s children has only the potential to make parents, and perhaps their children, better off because it enlarges the scope of choices available to parents. In standard single-person decision theory, the usual ar- gument is that more choice, rather than less, renders an individual decision-maker better off. The reasoning is that if one of the addi- tional choices is selected, then the individual has demonstrated, by making that choice, that she is better off than she would be without that option. If none of the additional alternatives are chosen, then the additional choices may simply be ignored. The “revealed preference” argument, however, ignores the pos- sibility that choice itself can create anxiety, concern, and regret about procedural or substantive issues related to choosing. The de- sire to avoid such feelings of anxiety or the responsibility of choice, as well as the costs of making fully-informed choices, provides alter- ect, in THE CODE OF CODES: SCIENTIFIC AND SOCIAL ISSUES IN THE HUMAN GENOME PROJECT 281, 298-99 (Daniel J. Kevles & Leroy Hood eds., 1992) (dis- cussing the inherent ambiguity of what is considered “normal” and its cultural and socially constructed aspects). 15. See Faye Flam, Designing the Family Tree a Road to Eugenics?, BUFF. NEWS, June 25, 1995, at F7 (describing a dwarf couple’s desire to genetically test their child for achondroplasia, the most common form of dwarfism). 16. See, e.g., JOHN A. ROBERTSON, CHILDREN OF CHOICE: FREEDOM AND THE NEW REPRODUCTIVE TECHNOLOGIES 170-71 (1994) (discussing intentional di- minishment); John A. Robertson, Genetic Selection of Offspring Characteristics, 76 B.U. L. REV. 421, 465-68 (1996) (discussing intentional diminishment from the perspective of parents’ rights to procreative liberty). 17. See generally MEHLMAN & BOTKIN, supra note 5 (discussing various eq- uity concerns raised by genetic technology). 18. Robert Wright, Achilles’ Helix, THE NEW REPUBLIC, July 9 & 16, 1990, at 27 (describing “homemade eugenics” as “individual families deciding what kinds of kids they want to have”). 19. THOMAS C. SCHELLING, MICROMOTIVES AND MACROBEHAVIOR 197 (1978) (discussing the possibility of choosing a child’s sex). W06-HUANG-WITH FIGURE 04/07/01 6:36 PM 1999] DESIGNER GENES 643 native rationales for individuals to use rules, standards, delegates, randomizing devices, heuristics, or other second-order decision mechanisms.20 In addition, the traditional argument that the alter- natives not chosen may simply be ignored also ignores the non- consequentialist value of having flexibility or options, even if those options are not used.21 Finally, it is well known that having fewer choices available might be advantageous in strategic interaction, es- pecially in bargaining.22 This game-theoretic value of reduced op- tions has applications in the designer genes context, as demon- strated in the Appendix. There are two additional problems with individual choice that are unique to individual free market reprogenetics. The first is that certain individuals who know that some of their traits are the result of their parents’ reprogenetic decisions may come to resent those choices. These individuals may come to wonder who they might have been, and care less for who they actually are, when those choices, such as a child’s sex, are irreversible or reversible only at high cost. As in other contexts, the phrase “living up to one’s poten- tial” might prove a heavy burden. This is even problematic if the reprogenetic technology is only successful with a probability less than one. Schelling discussed both possibilities and the clever example of how a couple’s decision regarding a third child’s sex may be viewed as their verdict on gender if that couple already had one child of each sex.23 The more general potential source of inefficiency here is that parents make reprogenetic decisions to maximize their own utility, instead of the utility of their unborn children. In a sense, this is the well-known conflict of interest problem that arises in any principal-agent relationship. The fact that the principal is unborn when the agent makes the reprogenetic decision is akin to intergenerational conflicts that might arise in environmental con- texts. A new wrinkle is that parents influence their children’s prefer- ences both consciously and subconsciously. Of course, if parents make harmful choices with regard to their children, and a third 20. See, e.g., CASS R. SUNSTEIN & EDNA ULLMAN-MARGALIT, SECOND-ORDER DECISIONS (University of Chicago Law School John M. Olin Law and Economics Working Paper No. 57, 2d Series 1998) (discussing various ways of making de- cisions). 21. See Kenneth J. Arrow, A Note on Freedom and Flexibility, in CHOICE, WELFARE, AND DEVELOPMENT: A FESTCHRIFT IN HONOR OF AMARTYA K. SEN 7-16 (Kaushik Basu et al. eds., 1995) (providing a formal model of the value of choices when preferences are flexible). 22. See AVINASH DIXIT & BARRY NALEBUFF, THINKING STRATEGICALLY: THE COMPETITIVE EDGE IN BUSINESS, POLITICS, AND EVERYDAY LIFE 152-55 (1991) (discussing historical examples of the strategic value of burning one’s bridges). 23. SCHELLING, supra note 19, at 209-10 (discussing some consequences of having a choice of a baby’s sex). W06-HUANG-WITH FIGURE 04/07/01 6:36 PM 644 WAKE FOREST LAW REVIEW [Vol. 34 party can verify that harm, the state may intervene on the chil- dren’s behalf. This policy, however, covers only harm to children once they are born. The wrinkle in the designer gene context is that such harm would not be apparent until the future, and is perhaps only probable. The California Supreme Court stated, in dictum, that if medical professionals warn parents that their child would be seri- ously ill, and the parents opt to proceed with the delivery, the child has a cause of action for wrongful life against the parents.24 The California legislature then responded by enacting legislation stating that “[n]o cause of action arises against a parent of a child based upon the claim that the child should not have been conceived or, if conceived, should not have been allowed to have been born alive.”25 While wrongful life lawsuits raise interesting ELSI26 and philo- sophical concerns,27 this Article focuses on another concern with re- progenetic choice: namely, the possible herd behavior that can result as the aggregate outcome of uncoordinated individual choices. Herd behavior in gene selection occurs when the utility that parents enjoy from a reprogenetic choice depends on the number of other parents who also make the same reprogenetic choice. This Ar- ticle focuses on the demand for, as opposed to the supply of, repro- genetic technologies.28 The supply of knowledge underlying repro- genetic technologies is ensured by the incentives of researchers to study and understand human genetics. The supply of reprogenetic technology is ensured by the profit motive. However, because it is impossible to put the gene back in the bottle, society can only learn how to best live with these new reprogenetic technologies. The rest of this Article is organized as follows. Section I dis- cusses possible sources of herd behavior for parental demand for re- progenetic technologies. Section II proposes a potential role for law 24. See Curlender v. Bio-Science Lab., 165 Cal. Rptr. 477, 488 (1980). 25. CAL. CIV. CODE § 43.6(a) (Deering 1994). 26. See, e.g., Alexander Morgan Capron, Tort Liability and Genetic Coun- seling, 79 COLUM. L. REV. 618, 661-66 (1979) (arguing that parents confronting a genetic risk should not face tort liability for deciding to conceive); Alexander Morgan Capron, Informed Decisionmaking in Genetic Counseling: A Dissent to the “Wrongful Life” Debate, 48 IND. L.J. 581, 603 (1973) (concluding that a child who suffers a genetic disease does not have a claim against the parents that de- cided to give birth despite the risks of the disease). But see, e.g., Alexander Morgan Capron, Which Ills to Bear?: Reevaluating the “Threat” of Modern Ge- netics, 39 EMORY L.J. 665, 695 (1990) (stating that his position has to be inevi- tably reconsidered upon biotechnological progress and social pressures). 27. See, e.g., Seana Valentine Shiffrin, Wrongful Life, Procreative Responsi- bility, and the Significance of Harm, 5 LEGAL THEORY 117, 148 (1999) (arguing for a more equivocal position regarding procreation). 28. See Eric A. Posner & Richard A. Posner, The Demand for Human Clon- ing, in FACTS AND FANTASIES ABOUT HUMAN CLONING 233 (Martha C. Nussbaum & Cass R. Sunstein eds., 1998) (explaining that, if there were no demand or a slight demand for such procedures as human cloning, then society would have no reason to fear, and therefore regulate, the practice of reprogenetic technolo- gies in order to prevent catastrophes). W06-HUANG-WITH FIGURE 04/07/01 6:36 PM 1999] DESIGNER GENES 645 and public policy in the effort to improve matters: namely, to provide a focal point for parents’ concerns about whether other parents will choose to use reprogenetic technologies for particular traits, and to thereby help coordinate individuals’ preferences regarding their children’s genetic traits. Section III discusses other concerns that arise in a world of designer genes. The Appendix provides two for- mal game-theoretical models of herd behavior when parents’ desires for particular genetic traits depend on the proportion of the popula- tion that also chooses those genetic traits for their children. Game theory permits formal analysis of such a “keeping up with the Jone- ses” phenomenon. I. SOURCES OF HERD BEHAVIOR FROM INDIVIDUAL REPROGENETIC CHOICE Herd behavior occurs when individual decision-makers imitate other decision-makers’ choices. Such lemming-like behavior may or may not be a good thing, both from the viewpoint of individuals and society. A social welfare analysis of herding depends on the source of that behavior. There are at least five possible reasons for herd- ing: learning from the behavior of others, the status quo bias, net- work effects, and positional goods or status competition with and without significant positive social externalities. These are depicted schematically in Figure 1. This Article will consider them in order. Figure 1: Sources of Herd Behavior Imitation or Learning Status Quo Bias Network Effects Positional Goods Behavior Signals Default Choice Popularity Status Competition Private Information Frame of Reference e.g. Right-handedness Winner-Take-All Significant Positive Insignificant Positive Social Externalities Social Externalities e.g. Intelligence Arms Races Self-Limiting Not Self-Limiting e.g. Blue-eyes e.g. Height W06-HUANG-WITH FIGURE 04/07/01 6:36 PM 646 WAKE FOREST LAW REVIEW [Vol. 34 First, decision-makers may have access to different information. If individuals are overly confident in their judgments,29 then moni- toring the actions of others could be desirable from both their own and society’s perspectives.30 Herding avoids duplicative information gathering or other decision-making costs. On the other hand, by simply mimicking the behavior of others, individuals and society may be worse off than if those people utilized their own private sub- stantive information.31 When behavior is considered a signal of pri- vate information, an availability cascade may result, whereby a be- lief is collectively formed in a self-reinforcing process, which gives that belief increased plausibility because of increased availability in public discourse.32 Herding due to informational or availability cas- cades, however, is likely not yet a concern in the designer genes con- text. The economist John Maynard Keynes, in stating that “[w]orldly wisdom teaches that it is better to fail conventionally than to suc- ceed unconventionally,”33 suggested that professional money man- agers may rationally practice investing with the herd to enhance reputations for investment abilities. Recent evidence suggests that younger fund managers follow the herd more than older fund man- agers because the younger ones are more likely than older ones to lose their jobs if their funds underperform the market.34 Similar 29. See Terrance Odean, Volume, Volatility, Price, and Profit When All Traders Are Above Average, 53 J. FIN. 1887, 1887 (1998) (demonstrating that overconfidence reduces the expected utility of those who are overconfident); Anne Kates Smith, A Little Investment Horse Sense: Overconfidence, Not the Lack of Data, Is the Biggest Danger, U.S. NEWS & WORLD REPORT, June 28, 1999, at 76 (reporting on empirical studies documenting overconfidence among horse race handicappers, discount-brokerage traders, and mutual fund inves- tors). 30. See GARY BELSKY & THOMAS GILOVICH, WHY SMART PEOPLE MAKE BIG MONEY MISTAKES AND HOW TO CORRECT THEM: LESSONS FROM THE NEW SCIENCE OF BEHAVIORAL ECONOMICS 176 (1999) (discussing the possibility that overconfi- dent people might overestimate their abilities to figure out what others think or know). 31. See Andrew Daughety & Jennifer Reinganum, Stampede to Judgment: Persuasive Influence and Herding Behavior by Courts, 1 AM. ECON. REV. (forth- coming 1999); see also Christopher Avery & Peter Zemsky, Multidimensional Uncertainty and Herd Behavior in Financial Markets, 88 AM. ECON. REV. 724, 728-30 (1998) (demonstrating the impossibility of informational cascades and herd behavior in financial markets when there is only a single dimension of risk). But see Eric Talley, Precedential Cascades: A Critical Appraisal, Olin Working Paper 99-6, U.S.C. Law School (1999), S. CAL. L. REV. (forthcoming Nov. 1999) (critically analyzing an informational cascade theory of legal prece- dent). 32. See Timur Kuran & Cass R. Sunstein, Availability Cascades and Risk Regulation, 51 STAN. L. REV. 683, 683 (1999) (defining “availability cascade”). 33. JOHN MAYNARD KEYNES, THE GENERAL THEORY OF EMPLOYMENT INTEREST AND MONEY 158 (1936) (expressing skepticism over the capability and will of long-term investors to engage in contrarian investment strategies). 34. See Judith Chevalier & Glenn Ellison, Career Concerns of Mutual Fund W06-HUANG-WITH FIGURE 04/07/01 6:36 PM 1999] DESIGNER GENES 647 herd behavior and motivations could apply to decisions by banks to lend money to developing countries,35 investments by corporate managers,36 information gathering by short-term speculators,37 and trading strategies by hedge funds and trading departments of in- vestment banks.38 In addition to such financial examples of herd- ing, law students often herd in choosing legal courses, areas of prac- tice, and summer or permanent associate positions at corporate firms. If later herd members differ in their tastes from earlier herd members, but are not aware of this difference, then later herd mem- bers are taking actions that maximize the preferences of earlier herd members instead of their own preferences. If preferences are malle- able or culturally constructed, so that what is chosen becomes that which is desired, then it may be difficult for herd members or ana- lysts to determine whether later herd members are better or worse off than they would have been in the absence of herding. The latter phenomenon can arise with certain traits, as modeled in the Appen- dix. A second motivation for herding is the well-documented status- quo bias, which suggests that benchmarks matter for choices.39 The status-quo bias is a bias toward default contract terms.40 In times of changing social norms, the fact that preferences are often context- dependent means that what is considered the default option may ef- fectively become mandatory. One way to avoid status-quo bias is to Managers, 114 QUART. J. ECON. 389, 409-16, 420, 430 (1999) (finding that, in a study of 453 fund managers, younger fund managers herd into more conven- tional portfolios and more popular sectors than older fund managers); Darren McDermott, Young Managers Follow the Herd, WALL ST. J., Aug. 2, 1999, at C1, C25 (reporting on studies that compared the investing practices of younger and older fund managers). 35. See SAMUEL GWYNNE, SELLING MONEY 58 (1986) (discussing the per- verse incentives that a credit analyst faces in assessing country risk). 36. See David S. Scharfstein & Jeremy C. Stein, Herd Behavior and In- vestment, 80 AM. ECON. REV. 465, 465 (1990) (demonstrating that corporate managers can exhibit herd behavior in their investment decisions when they are concerned about their reputations in the labor market). 37. See Kenneth A. Froot et al., Herd on the Street: Informational Ineffi- ciencies in a Market with Short-Term Speculation, 47 J. FIN. 1461, 1472 (1992) (demonstrating that short-term speculators may herd on the same particular sources of information that other informed traders know information completely unrelated to market fundamentals). 38. See Franklin R. Edwards, Hedge Funds and the Collapse of Long-Term Capital Management, 13 J. ECON. PERSP. 189, 206 (1999) (discussing the so- called “copycat” problem of many hedge funds and investment banks that have similar positions in order to make it difficult for all of them to liquidate their positions simultaneously). 39. See William Samuelson & Richard Zeckhauser, Status Quo Bias in De- cision Making, 1 J. RISK & UNCERTAINTY 7, 8 (1988) (providing experiments that demonstrate the “status-quo bias”). 40. See Russell Korobkin, The Status Quo Bias and Contract Default Rules, 83 CORNELL L. REV. 608, 608 (1998) (arguing that default rules in contract law may be sticky). W06-HUANG-WITH FIGURE 04/07/01 6:36 PM 648 WAKE FOREST LAW REVIEW [Vol. 34 prevent actors from knowing the status quo, but such may not be possible in the context of designer genes. A third rationale for herding is the presence of network exter- nalities.41 The concept of network externalities originated in the context of computer hardware, software, and telecommunications modes. In that context, a network externality occurs if the payoff or utility in choosing a particular type of computer hardware, software, or telecommunications mode is affected by the number of other users of the identical technology. For example, the utility of a particular type of computer hardware or software conceivably increases with the number of people who also choose that hardware or software be- cause of increased consumer support, increased ability to exchange or share information with others, and increased resale to or help from other users. Similarly, the benefits of being part of the tele- phone system or the Internet increase with the number of people on that system. Network externalities have been applied to contract terms and corporate law.42 Network externalities result in the po- tential for inefficiency that is, as with any type of externality, due to a conflict between individual rationality versus social optimality. Of course, as with many other cases of legal applications of economic ideas, the range of possible applications of network externalities to law may have been overstated.43 In addition, in all of the above situations, there may be negative network externalities that result from congestion or delay if the herd size exceeds network capacity. Examples of traits possibly exhibiting positive network exter- nalities are right-handedness, due to the overwhelming availability of products for right-handed people, or a “language gene,” which would facilitate the learning of multiple languages. Both situations are statically unproblematic if it is optimal for individuals to have those genes, but may be dynamically problematic due to lock-in ef- fects from a large installed base should things change over time. The flip side of the above discussion is the fear that children with those genetic traits will be disfavored by society at large, which re- sults in no children with those genetic traits. This potential all-or- nothing problem is related to the “tipping” phenomenon discussed in a famous book by Thomas Schelling.44 Tipping refers to how aggre- 41. See HAL R. VARIAN, INTERMEDIATE MICROECONOMICS: A MODERN APPROACH 591-97 (4th ed. 1996); see also Michael L. Katz & Carl Shapiro, Sys- tems Competition and Network Effects, 8 J. ECON. PERSP. 93, 97 (1994) (provid- ing expositions of network externalities). 42. See Michael Klausner, Corporations, Contracts, and Networks of Con- tracts, 81 VA. L. REV. 757, 789-840 (1995) (providing applications network ex- ternalities to corporate contracts). 43. See Mark A. Lemley & David McGowan, Legal Implications of Network Economic Effects, 86 CA. L. REV. 479, 590 (1998) (assessing the appropriateness of applying network externalities to legal settings). 44. SCHELLING, supra note 19, at 204-05 (discussing tipping as a cultural consequence of selection).