A Quiet Revolution in Welfare Economics- by Michael Albert and Robin Hahnel




Lionel Robbins, An Essay on the Nature and Significance of Economic Science, 2nd ed. (London: Macmillan, 1952).

1. John Stuart Mill, "Of the Liberty of Thought and Discussion," in The Philosophy of John Stuart Mill, Marshall Cohen, ed. (New York: modern Library, 1961),204-206.

2. See Donald E. Campbell, Resource Allocation Mechanisms (Cambridge: Cambridge University Press, 1987) chap. 4, for a concise, self-contained presentation of recent successes in extending welfare theory to an uncertain world. In his words on page 65, "Supposing that individual preferences do not change as random events are realized ... the Arrow-Debreu interpretation of the market mechanism is easily extended to accommodate uncertainty. All that is required is a reinterpretation of the notion of a commodity. Just as we distinguish an automobile delivered at one date from the same physical object at another date, insisting that they be identified as two different goods, so we also distinguish between an automobile delivered at one date when one specific random event is realized and the same automobile available at the same date under a different realization. There will be two different commodities and usually two different prices. Under this reinterpretation, an equilibrium of the market mechanism is Pareto optimal provided that a market exists for each of the newly defined goods."

Campbell goes on to explain on pages 65-66 that while the "contingent contract" extension of the Arrow-Debreu model "would require an intolerably large number of markets in practice," and that "an individual would spend all his time negotiating contracts in an astronomical number of markets," a "new mechanism" that includes "markets for securities, one for each possible random event," guarantees that "Pareto-optimal alloca tions once again emerge at equilibrium," although "the completeness of markets condition is still suspect."

We quote at such length because we are well aware that much recent work in welfare economics has been concerned with the implications of uncertainty, yet we ignore this issue and literature completely. We feel justified in doing so because the "formal solution" of expanding the list of goods for which contingent contracts are written, and the "practical solution" of competitive insurance markets, resolve the issue at the theoretical level with few surprises, as Campbell explains, provided one abstracts from the complications of preference change and incentive compatibility-two issues we treat in great depth.

3. For example, K. Arrow, "General Economic Equilibrium: Purpose, Analytic Techniques, Collective Choice," American Economic Review 64, no. 3 (June 1974): 253-72; and J. Hirshleifer and John G. Riley, "The Analytics of Uncertainty and Information-An Expository Survey," Journal of Economic Literature 17 (December 1979): 1375-142 1.

4. For example, K. Arrow and F. Hahn, General Competitive Analysis (San Francisco: Holden-Day, 1971); Peter Diamond and Michael Rothschild, eds., Uncertainty in Economics: Readings and Exercises (New York: Academic Press, 1978); and Jean Tirole, The Theory of Industrial Organization (Cambridge, Mass.: NUT Press, 1988).

5. For example, Robert Pollak, "Changes in Consumer Preferences: Endogenous Tastes in Demand and Welfare Analysis," American Economic Review 68, no. 2 (May 1978): 374-79.

6. It was Adam Smith's genius to know where to locate the fence without a map showing oil deposits and without knowing what "traditional" technology would become.

7. Leonid Hurwicz, "The Design of Mechanisms for Resource Allocation," in Frontiers of Quantitative Economics, Michael D. Intriligator and D. A. Kendrick, eds. (Amsterdam: North Holland Publishing, 1974).


1. H. Sidgwick, The Methods of Ethics, 7th ed. (London: Macmillan, 1907), as excerpted in Economic Justice, E. S. Phelps, ed. (Baltimore: Penguin, 1973),227.

2. Ibid.

3. John Rawls, A Theory of Justice (Cambridge, Mass.: Harvard University Press, 1971), 26.

4. John Stuart Mill, On Bentham and Coleridge (New York: Harper Torchbook, 1962), 94.

5. Jacob Viner, quoted in Herb Gintis, "Alienation and Power: Towards a Radical Welfare Economics" (Ph.D. diss., Harvard University, May 1969), 151.

6. 1. M. D. Little, A Critique of Welfare Economics, 2nd ed. (Glasgow: Oxford University Press, 1950), 258.

7. Lionel Robbins, The Nature and Significance of Economic Science (London: Macmillan, 1932),24.

8. Gintis, "Alienation and Power," 110.

9. Little, Critique, 276.

10. K. Arrow, "General Economic Equilibrium: Purpose, Analytic Techniques, Collective Choice," American Economic Review 64, no. 3 (June 1974): 253-72.

11. Since a careful analysis and criticism of "traditional assumptions" is a primary subject of the remaining chapters in part 1, we will not dwell on them here. But we wish to note that "traditional assumptions" might prove little more than a euphemism for "whatever assumptions are necessary to render the desired conclusions."

12. T. C. Koopmans, Three Essays on the State of Economic Science (New York: McGraw Hill, 1957), 54.

13. K. Arrow, "Political and Economic Evaluation of Social Effects and Externalities," in Frontiers of Quantitative Economics, Michael D. Intrilligator and D. A. Kendrick, eds. (Amsterdam: North Holland Publishing, 1974),6.

14. All references will be to John Rawls, A Theory ofJustice (Cambridge, Mass.: Harvard University Press, 1971); and Robert Nozick, Anarchy, State, and Utopia (New York: Basic Books, 1974).

15. Rawls, Theory of Justice, 27-30.

16. Nozick, Anarchy, State, and Utopia, 32.

17. Ibid., 33.

18. See Nozick, Anarchy, State, and Utopia, 150-53, for an exposition of the entitlement approach to distribution, as opposed to the difference principle; and Rex Martin, Rawls and Rights (Lawrence, Kan.: University of Kansas Press, 1985), chap. 8. We do not treat this disagreement in depth because it is peripheral to our concerns. But it is worth noting that the difference basically divides the politically liberal new contractarians (Rawlsians) from the politically conservative new contractarians (Nozickians), the latter group dominating among adherents to the contractarian banners in the economics profession.

19. In private communication Rawls has indicated he does not exclude the possibility that public enterprise, competitive market institutions might also be compatible with his liberty and maximin principles.

20. These developments are presented in a variety of advanced microeconomic theory texts, but Gerard Debreu, Theory of Value (New York: John Wiley and Sons, 1959), is the original culmination of the modern, choice theoretic version of neoclassical welfare theory.

21. The "principle" of Occam's razor is, of course, that if the same set of conclusions can be deduced from a reduced set of assumptions, the reduction is to be preferred. The loss of the "bliss point" from the set of conclusions nullifies the application of the principle in this case.

22. Rawls, Theory of Justice, 23, 24, 27, 29, 30.

23. D. M. Winch, Analytical Foundations of Welfare Economics (Baltimore: Penguin, 1973), 144.

24. See Winch, Ibid., 135-50; or Yew-Kwang Ng, Welfare Economics, rev. ed. (London: Macmillan, 1983), chap. 3, for a concise presentation of the different forms of compensation tests and their contradictoriness.

25. See Milton Friedman, "The Distribution of Income," in Economics: Mainstream Readings and Radical Critiques, 3rd ed., David Mermelstein, ed. (New York: Random House, 1976), for an essay that presents this argument concisely.

26. Rawls, Theory of Justice, 27.

27. Ibid., 28-29.

28. Nozick, Anarchy, State, and Utopia, 3 1.

29. Rawls, Theory of Justice, 27-28.

30. Ibid., 29.

3 1. Ibid., 27.

32. Ibid., 23.

33. While the ability of centrally planned and market public enterprise economies to achieve a full range of Pareto optimal outcomes, in theory, under assumptions no more generous than those granted private enterprise market economies, is not unknown, it forms a large part of what we call the "old debate," and many still challenge this conclusion. For this reason we begin our treatments of markets, ownership, and central planning in part 3 by clarifying the "old debate" before proceeding to more important concerns.

34. Rawls, Theory of Justice, 29.

35. Ibid.

36. While it is obvious we feel a contractarian framework best expresses the traditional welfare paradigm, this is a peripheral point and irrelevant to the rest of our argument. For our principal quarrel is with the traditional paradigm itself, not with its form of expression.

37. John Stuart Mill, in The Philosophy of John Stuart Mill, Marshall Cohen, ed. (New York: modern Library, 1961).

38. It is of interest that in all three cases the ways in which each variant of traditional welfare theory seeks to protect the sanctity of the individual produce as a by-product an implicit denial of the endogeneity of preferences. In chapter 4 we will review Mill's discussion of how classical utilitarians' denial of a hierarchy of pleasures prevented them from analyzing the phenomenon of preference development, and Gintis' argument for how the Robbins Principle blinds neoclassical theorists to the welfare effects of endogenous preferences. Our own argument that the new contractarian treatment of the original position entails an implicit denial of the endogeneity of preferences as well is presented in chapter 6.

39. Nozick, Anarchy, State, and Utopia, 32-33.

40. Rawls, Theory of Justice, 29.


1. See Paul A. Samuelson, Foundations of Economic Analysis (Cambridge, Mass.: Harvard University Press, 1947), 57; or C. E. Ferguson, The Neoclassical Theory of Production (Cambridge: Cambridge University Press, 1971), 7.

2. See Gerard Debreu, Theory of Value (New York: Wiley, 1959); or K. Arrow and F. H. Hahn, General Competitive Analysis (San Francisco: Holden-Day, 197 1), chap. 3.

3. While abstracting from the production process no doubt permitted sharper focus on many aspects of the logic of exchange relations, it always ran the risk of misunderstanding phenomena that are the product of interrelations between the logics of exchange and production processes. While this may appear obvious, once the "abstraction" was embodied in the traditional paradigm most economists lost track of it.

4. See: Harry Braverman, Labor and Monopoly Capital (New York: Monthly Review Press, 1974); Michael Burawoy, Manufacturing Consent (Chicago: University of Chicago Press, 1979); Richard Edwards, Contested Terrain (New York: Basic Books, 1979); David Gordon, Richard Edwards, and Michael Reich, Segmented Work, Divided Workers (Cambridge: Cambridge University Press, 1982); David Noble, America By Design (New York- Knopf, 1977); Andrew Zimbalist, Case Studies in the Labor Process (New York: Monthly Review Press, 1979); Stephen Marglin, "What Do Bosses Do?," Review of Radical Political Economics

6, no. 2 (Summer 1974); William Lazonick, "The Subjugation of Labor to Capital: The Rise of the Capitalist System," Review of Radical Political Economics 10, no. 1 (Spring 1978); Michael Reich and James Devine, "The Microeconomics of Conflict and Hierarchy in Capitalist Production," Review of Radical Political Economics 12, no. 4 (Winter 1981); Kathy Stone, "The Origins of Job Structures in the Steel Industry," Review of Radical Political Economics 6, no. 2 (Summer 1974); Michele Naples, "The Structure of Industrial Relations, Labor Militancy and the Rate of Growth of Productivity: The Case of U.S. Mining and Manufacturing, 1953-1977" (Ph.D. diss., University of Massachusetts at Amherst, 1982); Herb Gintis, "The Nature of the Labor Exchange," Review of Radical Political Economics 8, no. 2 (Summer 1976); Samuel Bowles, "The Production Process in a Competitive Economy: Walrasian, Neo-Hobbesian, and Marxian Models," American Economic Review 75, no. I (March 1985).

5. P. Doeringer and M. Piore, Internal Labor Markets and Manpower Analysis (Lexington, Mass.: D. C, Heath, 1971).

6. Armen Alchian and Harold Demsetz, "Production Information Costs and Economic Organization," American Economic Review 62, no. 5 (1972); Guillermo Calvo, "Quasi Walrasian Theories of Unemployment," American Economic Review 69, no. 2 (May 1979); Edward Lazear, "Agency, Earnings Profiles, Productivity, and Hours Restrictions," American Economic Review 7 1, no. 4 (September 198 1); Oliver Williamson, "The Organization of Work," Journal of Economic Behavior and Organization, no. 1 (1980).

7. See our discussion in chapters 2 and 4 of Unorthodox Marxism (Boston: South End Press, 1978) to the effect that Marx's Labor Theory of Value is not only unnecessary to explain the origins of profits, but confuses matters considerably. For a cogent, albeit polemical, argument for a Sraffian theory of prices, wages, and profits as opposed to a Marxian theory see Ian Steedman, Marx After Sraffa (London: New Left Books, 1977). For a more rigorous demonstration of the sufficiency of a Sraffian theory of prices wages, rents, and profits in the more general context of joint production, fixed capital, land, and natural resources see Gilbert Abraham-Frois and Edmond Berrebi, Theory of Value Prices, and Accumulation (Cambridge: Cambridge University Press, 1979), chap. 3. For a demonstration that generalizing the production environment from linear technologies to convex production sets still permits an explanation of profits as due to "exploitation" in terms other than labor values, but renders an explanation of exploitation and profits via the Labor Theory of Value impossible, see John Roemer, Analytical Foundations of Marxian Economic Theory (Cambridge: Cambridge University Press, 1981), chap. 2. Finally, for a highly provocative and ingenious treatment of exploitation in general and misconceptions in the Marxist theory of exploitation, see John Roemer, A General Theory of Exploitation and Class (Cambridge, Mass.: Harvard University Press, 1981); and John Roemer, Free to Lose (Cambridge, Mass.: Harvard University Press, 1988).

8. Prior to the publication of Cornelius Castoriadis, Political and Social Writings, edited and translated by David Ames Curtis (Minneapolis: University of Minnesota Press, 1988), only a few pamphlets published under the pseudonym of Paul Cardan in the 1960s and some essays in Telos in the 1970s were available in English. For an excellent review of Castoriadis' work on the occasion of its long overdue printing in the U.S. see Carl Boggs' review in Z Magazine (January 1989): 75-77.

9. The essay was finally published long after stirring up a more significant controversy than most published articles in the Review of Radical Political Economics 6, no. 2 (Summer 1974).

10. Stone, "Job Structures in the Steel Industry."

11. For a sample of work responding to Braverman's efforts see Politics and Society 8, nos. 3, 4 (1978) devoted exclusively to consideration of his work.

12. Herb Gintis, "The Nature of the Labor Exchange and the Theory of Capitalist Production," Review of Radical Political Economics 8, no. 2 (Summer 1976).

13. Herb Gintis, "The Nature of the Labor Exchange: Toward a Radical Theory of the Firm," Harvard Institute of Economic Research, discussion paper no. 328 (October 1973): 1-2.

14. Herb Gintis and Donald Katzner, "Profits, Optimality, and the Social Division of Labor in the Finn," in Sociological Economics, M. Levy-Garboua, ed. (London: Sage, 1978), chap. 9.

15. R. Coase, "The Nature of the Finn," Economica 4 (November 1937). 16. Herbert Simon, Models of Man (New York: Wiley, 1957).

17. Coase, "Nature of the Firm," 333.

18. Simon, Models of Man chap. 11.

19. Gintis and Katzner, "Profits," 270.

20. Ibid.

2 1. Gintis, "Labor Exchange: Toward a Radical Theory," 18.

22. Gintis and Katzner, "Profits," 289.

23. Samuel Bowles, "The Production Process in a Competitive Economy: Walrasian, Neo-Hobbesian, and Marxian Models," American Economic Review 75, no. I (March 1985): 16-36. For related work which draws less "radical" conclusions see B. Eaton and W. White, "Agent Compensation

and the Limits of Bonding," Economic Inquiry 20, no. 3 (July 1982): 330-42; J. Malcomson, "Unemployment and the Efficiency Wage Hypothesis," Economic Journal 91 (December 1981): 848-66; C. Shapiro and J. Stiglitz, "Equilibrium Unemployment as a Worker Discipline Device," American Economic Review 74, no. 3 (June 1984): 433-44; and R. Sparks, "A Model of Involuntary Unemployment and Wage Rigidity: Worker Incentive and the Threat of Dismissal," Journal of Labor Economics 4, no. 4 (1986): 560-8 1. A useful summary of treatments along traditional lines can be found in J. Yellen, "Efficiency Wage Models of Unemployment," American Economic Review 74, no. 2 (1984): 200-205.

24. Bowles, "Production Process," 22.

25. Ibid., 32.

26. Ibid., 33.

27. Ibid., 3 1.

28. For Bowles' views on this subject see Samuel Bowles, David Gordon, and Thomas Weisscopf, Beyond the Wasteland (Garden City, N.Y.: Anchor Doubleday, 1983).

29. Gil Skillman, "Microfoundations of the Conflict Theory of the Firm" (Ph.D. diss., University of Michigan, 1985).

30. S. Ross, "The Economic Theory of Agency: The Principal's Problem," American Economic Review 63, no. 2 (May 1973); idem, "On the Economic Theory of Agency and the Principal of Similarity," in Essays on Economic Behavior Under Uncertainty, M. Balch, D. McFadden, and S. Wu, eds. (Amsterdam: North Holland Publishing, 1974); J. A, Mirrlees, "NOTES on Welfare Economics, Information, and Uncertainty," in Balch, McFadden, and Wu, eds.; idem, "The Optimal Structure of Incentives and Authority within an Organization," Bell Journal of Economics 7, no. 1 (Spring 1976); M. Harris and A. Raviv, "Some Results on Incentive Contracts with Applications to Education and Employment, Health Insurance, and Law Enforcement," American Economic Review 68, no. 1 (March 1978); S. Shavell, "Risk Sharing and Incentives in the Principal and Agent Relationship," Bell Journal of Economics (Spring 1979); M. Harris and R. Townsend, "Resource Allocation Under Asymmetric Information," Econometrica 40, no. I (January 198 1); and S. Grossman and 0. Hart, "An Analysis of the Principal-Agent Problem," Econometrica 5 1, no. 1 (January 1983).

31. D. Sappington, "Limited Liability Contracts Between Principal and Agent," Journal of Economic Theory 29, no. I (Feb 1983).

32. J. Stiglitz, "Incentives, Risk, and Information: notes Towards a Theory of Hierarchy," Bell Journal of Economics 6, no. 2 (Autumn 1975).

33. B. Holmstrom, "Moral Hazard and Observability," Bell Journal of Economics (Spring 1979); D. Sappington, "Incentive Contracting with

Asymmetric and Imperfect Contractual Knowledge," Journal of Economic Theory 34. no. I (October 1984); M. Harris and A. Raviv, "Optimal Incentive Contracts with Imperfect Infonrmation," Journal of Economic Theory 20. no. 2 (April 1979); J. Christensen, "Communication in Agencies,"BellJournalofEconomics 12,no.2(Autumn 1981); and G. Skillman and H. Ryder, "Bargaining with Costly Unilateral Replacement," Department of Economics, Brown University, working paper no. 88-17 (June 1988).

34. For example: B. Holmstrom, "Moral Hazard in Teams," Bell Journal of Economics 13, no. 2 (Autumn 1982); E. Lazear and S. Rosen, "Rank Order Tournaments as Optimum Labor Contracts," Journal of Political Economy 89, no. 5 (1981); R. Radner, "Optimal Equilibrium in some Repeated Games with Imperfect Monitoring," Bell Laboratories discussion paper, May 1981; idem, "Monitoring Cooperative Agreements in a Repeated Principal-Agent Relationship," Bell Laboratories economic discussion paper no. 184 (March 1981); and G. Skillman, "Sequential Bargaining in Capitalist Firms," Department of Economics, Brown University, working paper no. 88-28 (November 1988).

35. For recent overviews of empirical work see F. Blau and M. Ferber, "Discrimination: Empirical Evidence for the U.S.," American Economic Review 77, no. 2 (1987): 316-20; J. Cotton, "On the Composition of Wage Differentials," Review of Economics and Statistics 70, no. 2 (1988): 236-43; G. Cain, "The Economic Analysis of Labor Market Discrimination: A Survey," in Handbook of Labor Economics, 0. Ashenfelter and R. Layard, eds. (New York: Elsevier Science Publications, 1986); and D. Neumark, "Employers' Discriminatory Behavior and the Estimation of Wage Discrimination," Journal of Human Resources 23, no. 3 (1988): 279-95.

36. Michael Reich, appendix to chap. 5, "Models of Class Conflict and Racial Inequality Within the Firm," Racial Inequality (Princeton, N.J.: Princeton University Press, 1981), 204-15. For similar theoretical treatments see also: Michael Reich and James Devine, "The Microeconomics of Conflict and Hierarchy in Capitalist Production," Review of Radical Political Economics 14, no. 4 (Winter 198 1); John Roemer, "Divide and Conquer: Microfoundations of the Marxian Theory of Wage Discrimination," Bell Journal of Economics 10, no. 2 (Autumn 1979); and Joel Kaplan, "The Divide and Conquer Hypothesis: An Analysis of Internal Labor Markets and Racial Inequality Among Men in the U.S." (Ph.D. diss., American University, 1985).

37. K. Arrow, "Models of Job Discrimination," in Racial Discrimination in Economic Life, A. Pascal, ed. (Lexington, Mass.: Lexington Books, 1972).

38. All quotations in this section are from Arrow, "Models of Job Discrimination," 83-99.

39. Since Arrow does not spell out all his assumptions, we cannot know whether or not he assumes employee immortality or the necessity of a continuous replacement process. In the case of mortals, it would appear the word "always" here may be too strong. And if employers' need to replace mortal workers eventually could be accomplished by infrequent but large scale replacements, the wage differentials should tend to disappear in what we might call the long, long run too. Perhaps more telling is the crucial role played by the assumption "there is only one kind of utility function expressing a trade-off between wages and the proportion of white workers." Just as "competition" only drove out employers' racism when we abandoned the assumption of uniform employer tastes for discrimination, if we allow that white employees are racist to different degrees, it seems wage differentials would eventually vanish as well.

40. The parallel to the Marxist theory of the tendency of the rate of profit to fall due to the introduction of "capital intensive" techniques of production is too striking to go without mention. Marx offered the theory with a list of similarly "counteracting tendencies." However, Okishio proved conclusively there is no need to consider counteracting tendencies since the rate of profit does not tend to fall for the reason Marx originally supposed. Or, for those who (erroneously) insist on continuing to think in terms of "the tendency" and "counteracting tendencies," the "counteracting tendencies" always outweigh the "tendency." Latter-day defenders of the Marxist faith tried to rescue the theory of the tendency for the rate of profit to fall by suggesting that it can still operate in models with fixed capital, or by modifying the usual investment criterion. All was to no avail except in the eyes of those who refuse to abandon a cherished part of their paradigm regardless of, in this case, proven theoretical internal inconsistencies. See John Roemer, Analytical Foundations of Marxian Economic Theory (Cambridge: Cambridge University Press, 1981), chaps. 4 and 5 for a rigorous treatment. The latest tack in the rescue attempt is strikingly similar to Arrow's departure from the standard assumption of the traditional debate. Where Arrow proposes "that we look more closely at the long run adjustment process" in context of nonconvexities rather than search for an alternative explanation that works under traditional neoclassical assumptions, Marxist diehards have tried to rescue one of their favorite convictions by abandoning comparative statics adjustment assumptions. Once one enters the murky waters of disequilibrium adjustment dynamics [uneven development] almost nothing can be established with certaintywhich in this case proves convenient for diehard believers in the Marxist tendency for the rate of profit to fall.

41. This is the assumption that contains the critical error. But without a paradigm and theory that analyzes the social dynamics that occur within

the production process, the assumption appears perfectly logical. When we look closely inside the black box of private enterprise production in chapter 8 we will see that this is not the case.

42. We do not mean to imply nothing is to be learned about economic discrimination from the neoclassical treatment. No doubt some of what Arrow has analyzed is going on. A full treatment of discrimination must encompass both possibilities-that white workers gain and that employers gain from discrimination against minorities-and analyze the circumstances that would or would not permit each from doing so. Our point is only that the traditional paradigm has blinded neoclassical theorists to a most important aspect of discrimination-the motivation for employers in competitive circumstances to discriminate.

43. Roemer, "Divide and Conquer"; and Reich and Devine, "Microeconomics of Conflict and Hierarchy."

44. Reich, "Models of Class Conflict and Racial Inequality Within the Firm," in Racial Inequality, 204-15 appendix to Chapter 5. We do not present Reich's model here because we develop a more general model that incorporates Reich's as a special case in chapter 8.

45. Joan Robinson, The Economics of Imperfect Competition (London: Macmillan, 1933), 301-304. For a contemporary application of the Robinson model to wage differences by sex see Janice Madden, "Discrimination: Manifestation of Male Market Power?," in Sex, Discrimination, and the Division of Labor, Cynthia Lloyd, ed. (New York: Columbia University Press, 1975), 146-74.

46. Doeringer and Piore, Internal Labor Markets. For Piore's methodological characterization of this work see M. Piore, "Labor Market Segmentation: To What Paradigm Does It Belong?," American Economic Review 73, no. 2 (May 1983): 249-53. For other recent contributions along institutionalist lines see G. Akerlof, "Gift Exchange and Efficiency Wage Theory: Four Views," American Economic Review 74, no. 2 (1984): 79-83; C. Fisher, "Towards a More Complete Understanding of Occupational Sex Discrimination, "Journal of Economic Issues 2 1, no. 1 (1987): 113-38; and S. Woodbury, "Power in the Labor Market: Institutionalist Approaches to Labor Problems," Journal of Economic Issues 2 1, no. 4 (1987): 1781-1807.

47. Glen Cain and Harold Watts, eds., Income Maintenance and Labor Supply: Econometric Studies (Chicago: Rand McNally College Publication, 1973), chap. 9.

48. One would have to examine what circumstances would generate such a condition. But seniority rules, firm-specific skills, and pension contribution/benefit structures are among a number of possibilities that have been explored.

49. Beyond blinding traditional theorists, the traditional paradigm has affected those who dared to hunt inside the forbidden demesne of the black box as well. While these poachers deserve credit for fearlessly pursuing their intuitions, the traditional paradigm tends to channel their hunting into the unequal information/uncertainty preserve that, if we are correct, is stocked with smaller game. This may even hold for principal-agent theory in its original domain of moral hazard and adverse selection, where future work may bag bigger game by moving to a multiperiod, game-theoretic approach where the human outputs of previous periods are known with ever greater accuracy by all parties and recognized as the human conditions that form the structure of comparative advantages in future conflicts-all of which should be taken into account in a theory of "rational choice."

50. Abba Lerner, "The Economics and Politics of Consumer Sovereignty," American Economic Review 62, no. 2 (May 1972): 259.


1. Richard Musgrave, The Theory of Public Finance (New York: McGraw Hill, 1959),6, 8.

2. Ibid., 10- 11.

3. P. A. Samuelson, "The Pure Theory of Public Expenditures," Review of Economics and Statistics 36, no. 4 (November 1954); idem, "Diagrammatic Exposition of a Theory of Public Expenditure," Review of Economics and Statistics 37, no. 4 (November 1955).

4. Musgrave, Theory of Public Finance, 8.

5. Ibid., 134.

6. Ibid.

7. Ibid., 135.

8. C. M. Tiebout, "A Pure Theory of Local Expenditures," Journal of Political Economy 64, no. 5 (October 1956): 416-23.

9. Musgrave, Theory of Public Finance, 132.

10. Ibid., 7-8.

11. See R. C. D'Arge and E. K. Hunt, "Environmental Pollution, Externalities and Conventional Economic Wisdom: A Critique," Environmental Affairs no. 1 (1971): 266-86. E. K. Hunt and R. C. D'Arge, "On Lemmings and Other Acquisitive Animals: Propositions on Consumption," Journal of Economic Issues 7, no. 2 (June 1973): 337-53; and E. K. Hunt, "A Radical Critique of Welfare Economics," in Growth, Profits, and Property, Ed Nell, ed. (Cambridge: Cambridge University Press, 1980), 239-49. James Campen, "Public Expenditure Analysis" (Ph.D. diss., Harvard

University Press, October 1976); idem, Benefit, Cost and Beyond (Cambridge, Mass.: Ballinger, 1986) to some extent echoes elements of Hunt's criticism in his own critique of traditional benefit cost analysis.

12. Hunt, "Radical Critique."

13. Ibid., 246.

14. Ibid., 245. Examination of determinants of the reproduction and transformation of the human and social characteristics that traditional welfare theory takes as "givens" is a constant theme in the alternative approach we favor. Here we find Hunt criticizing traditional theory for accepting "as somehow metaphysically given" the state of social interrelations that define the degree of external effects in the economy. In what follows Hunt outlines the incentives market systems establish for individuals to transform and manipulate what are traditionally taken as "givens."

15. Ibid., 245-46. Hunt goes on to point out that a symmetrical mechanism regarding positive external economies is precisely what is lacking in a free market system. "External economies also offer incentives for individual gain, but the incentive structure here is basically different than for external diseconomies. Without liability or nuisance rules that establish social responsibility, it is in the interest of both generator and recipient to negotiate on external diseconomies. However, with external economies the recipient gains more by attempting to be a free rider except, perhaps, at the margin. Consequently, the incentive for creating or producing external economies is less than that for external diseconomies, except perhaps for altruists" (246).

16. Ibid., 241.

17. In other words, if it were the case that a set of economic institutions yielded Pareto optimal outcomes, we would call those institutions socially efficient. In part 3 we go to great lengths to show this is less likely to be the case for private enterprise, competitive market institutions than commonly assumed. Nor is it the case for public enterprise market economies, nor centrally planned economies, although for reasons not usually emphasized. But our quarrel is with the particular institutions that number among their deficiencies the generation of socially inefficient outcomes. Our quarrel is not with the formalization of the notion of social efficiency known as the Pareto criterion.

18. Whether the notion of social efficiency formalized as Pareto optimality deserves as prominent a role in welfare theory as it has been accorded traditionally, is a different matter, as is the question of whether additional formalizations of social efficiency are needed because the Pareto criterion is insufficient in public choice environments. We might add that it is not clear to us that Hunt really means to challenge Pareto optimality as a useful formalization of the notion of social efficiency as he clearly appears to do

in the passage quoted above. But regardless of Hunt's view, there is a regrettable tendency among not a few radical economists to reject the concept of Pareto optimality altogether.

19. Hunt, "Radical Critique," 241.

20. Ibid., 243.

21. Ibid., 242.

22. Ibid., 243.

23. R. Lipsey and K. Lancaster, "The General Theory of the Second Best," Review of Economic Studies 24, no. 1 (1956-57): 11-32.

24. Hunt, "Radical Critique," 243.

25. One must take this statement in the spirit it is offered. It is no doubt possible to choose combinations of economic institutions and welfare criteria in which the theory of the second best would not hold. We mean to suggest that for any reasonably interesting welfare theory and any reasonably interesting economy the theory of the second best will likely apply. Certainly if we insist that our welfare theory include the Pareto criterion, the applicability of the theory of the second best cannot be sensitive to modifications in welfare approach.

26. We regret ending our treatment of Hunt's criticisms of neoclassical welfare theory on a critical note. In our view he is one of the most insightful, modern critics of neoclassical welfare theory. It is most regrettable that his analysis of externalities is known to so few mainstream and nomnainstream economists alike.

27. Jean-Claude Milleron, "Theory of Value with Public Goods: A Survey Article," Journal of Economic Theory 5, no. 3 (December 1972): 419-77.

28. Peter Hammond, "Symposium on Incentive Compatibility: Introduction," Review of Economic Studies 46, no. 2 (1979): 18 1.

29. Theodore Groves, "Efficient Collective Choice when Compensation Is Possible," Review of Economic Studies 46, no. 2 (1979).

30. W. Vickrey, "Counterspeculation, Auctions, and Competitive Sealed Tender," Journal of Finace 16, no. 1 (March 1961): 8-37.

3 1. Theodore Groves, "The Allocation of Resources under Uncertainty," (Ph.D. diss., University of California at Berkeley, 1970).

32. E. Clarke, "Multipart Pricing of Public Goods," Public Choice, 11 (1971):17-33.

33. L. Hurwicz, "On Informationally Decentralized Systems," in Decision and Organization, McGuire and Radner, eds., (Amsterdam: North Holland Publishing, 1972); and idem, "The Design of Mechanisms for Resource Allocation," American Economic Review 63 no. 2 (May 1973): 1-30.

34. J. H. Dreze and D. de la Vallee Poussin, "A Tatonnement Process for Public Goods," Review of Economic Studies 38 (1971): 133-50.

35. Ibid., 147.

36. Clark, "Multipart Pricing."

37. T. Groves and E. Loeb, "Incentives and Public Inputs," Journal of Public Economics, no. 4 (1975): 211-26.

38. T. Groves and J. Ledyard, "Optimal Allocation of Public Goods: A Solution to the 'Free Rider' Problem," Econometrica45, no. 4 (May 1977): 791-92.

39. Taking the two possible ways of lying in turn: if I understate my benefits I will have no direct effect on my own assessment and only very marginally increase the tax bills of all others. If I overstate my benefits I will, again, have no direct effect on my own assessment, and only very marginally decrease the tax bills of others. Since a budget balancing constraint implies that higher assessments for others will eventually yield a downward adjustment in my own assessment, a lingering incentive to understate one's benefits might appear to remain. But this effect is extremely secondary. It is like the effect that implies that a very aware participant in even a competitive market can improve upon "competitive" behavior by taking into account his or her minimal effect on market price-a point well known to traditional theorists.

40. Obviously, among millions of citizens I cannot expect to affect the outcome greatly. So for those who don't bother to vote in presidential elections because the probability of their affecting the outcome is insignificant, there may be no answer to the question why they should study options and report their marginal willingness to pay for public goods carefully. But the point is, there is no incentive to dissimulate. On the other hand, if I have information about the likely differences between my preferences for public goods and the preferences of the rest of my compatriots, there may be strong incentives to overexaggerate those differences when reporting to enhance my (limited) expected impact. But this is a problem that occurs in many contexts for which no "solution" has been found. We might note that as significant as incentive-compatible mechanisms certainly are, they shouldn't be expected to work miracles!

41. A. Gibbard, "Manipulation of Voting Schemes: A General Result," Econometrica 41, no. 4 (July 1973): 587-601; and M. Shatterthwaite, "Strategy-Proofness and Arrow's Conditions: Existence and Correspondence Theorems for Voting Procedures and Social Welfare Functions," Journal of Economic Theory 10, no. 2 (April 1975): 187-217.

42. J. Green and J. Laffont, "Revelation des Preferences pour les Biens Publics," Cahiers du Seminaire d'Econometrie (C.R.N.S., Paris); and L.

Hurwicz, "On the Existence of Allocation Systems Whose Manipulative

Nash Equilibria Are Pareto-Optimal" (Unpublished paper presented at 3rd World Congress of the Econometric Society, Toronto, 1975).

43. See Groves and Ledyard's "Quadratic Mechanism" in "Optimal Al

location," section 4; M. Walker's "Paired Difference Mechanism" in "An Informationally Efficient Auctioneerless Mechanism for Attaining Lindahl Allocations" (Unpublished paper); L. Hurwicz, "Shared Cost Mechanism" in "An Outcome Function with Lindahl Allocation at Nash Equilibria without an Auctioneer" (Unpublished paper); and idem, "Smooth Outcome Functions Yielding Walrasian Allocation at Nash Equilibria without an Auctioneer" (Unpublished notes).

44. Hunt, "Radical Critique," 246.


1. Milton Friedman, Price Theory (Chicago: Aldine Press, 1962), 13.

2. W. M. Gorman, "Tastes, Habits, and Choices," International Economic Review 8 no. 2 (June 1967): 218.

3. Jon Elster, Logic and Society (New York: Wiley, 1978), 39-40.

4. This theme will be further developed in chapter 8.

5. John Stuart Mill, "On Liberty," sec. 1, paragraph 11, in The Philosophy of John Stuart Mill, Marshall Cohen, ed. (New York: modern Library, 1966).

6. Bertrand Russell, Philosophical Essays, rev. (London: Allen and Unwin, 1966),21.

7. John Stuart Mill, in The Philosophy of John Stuart Mill, Cohen, ed., 331-32.

8. See Sidgwick as quoted in chap. 1, footnote 1.

9. Quoted in Herb Gintis, "Alienation and Power: Towards a radical Welfare Economics" (Ph.D. diss., Harvard University, May 1969),153.

10. Gintis, "Alienation and Power," 137.

11. Ibid., 137-3 8, 187.

12. Ibid., 184.

13. Ibid., 185-87.

14. Cited in Elster, Logic and Society, 40.

15. N. Georgescu-Roegen, "The Theory of Choice and the Constancy of Economic Laws," Quarterly Journal of Economics 64, no. I (February 1950): 125-38; R. Strotz, "Myopia and Inconsistency in Dynamic Utility Maximization," Review of Economic Studies 23 (1955-56):165-80; and J. Hasanyi, "Welfare Economics of Variable Tastes," Review of Economic Studies 21 (1954): 204-13.

16. C. C. von Weizsacker, "notes on Endogenous Change of Tastes," Journal of Economic Theory 3 no. 4 (December 1971): 345-72.

17. Peter Hammond, "Endogenous Tastes and Stable Long-Run Choice," Journal of Economic Theory 13 no. 2 (October 1976): 329-40; C. Lluch, "Expenditure, Savings and Habit Formation," International Economic Review 15 no. 3 (October 1974): 786-97; and Marcel Boyer, "A Habit Forming Optimal Growth Model," International Economic Review 19 no. 3 (October 1978): 585-609.

18. Robert Pollak: "Habit Formation and Dynamic Demand Functions," Journal of Political Economy 78 (July/August 1970): 745-63; idem, "Additive Utility Functions and Linear Engel Curves," Review of Economic Studies 38 (1971): 401-14; idem, "Habit Formation and Long-Run Utility Functions," Journal of Economic Theory 13 no. 2 (October 1976): 272-97; idem, "Changes in Consumer Preferences: Endogenous Tastes in Demand and Welfare Analysis," American Economic Review 68, no. 2 (May 1978): 37479; idem, "Welfare Evaluation and the Cost of Living Index in the Household Production Model," American Economic Review 68 no. 3 (June 1978): 28599; Robert Pollak and Terence Wales, "Estimation of Complete Demand Systems from Household Budget Data: The Linear and Quadratic Expenditure Systems," American Economic Review 68 no. 3 (June 1978): 348-59; and idem, "Equity: The Individual vs. the Family: Welfare Comparisons and Equivalence Scales," American Economic Review 69 no. 2 (May 1979): 2162 1.

19. Von Weizsacker, "Endogenous Change of Tastes," 345.

20. Ibid., 346.

2 1. Ibid.

22. Ibid., 347.

23. A. El-Safty, "Adaptive Behavior, Demand and Preferences," Journal of Economic Theory 13 no. 2 (October 1976): 298-318.

24. Ibid., 298.

25. Pollak, "Habit Formation," 296.

26. We say of little additional concern to welfare theory because the problems of nonconvexity of preferences have long been recognized. As von Weizsacker admits, his model was specifically excluded by Marshall in his formulations of the usual optimality conclusions of traditional welfare theory under what Marshall formulated as the assumption of decreasing marginal utility: "There is, however, an implicit condition in this law which should be made clear. It is that we do not suppose time to

be allowed for any alteration in the character or tastes of the man himself. It is, therefore, no exception to the law that the more good music a man hears, the stronger is his taste for it likely to become; that avarice and ambition are often insatiable; or that the virtue of cleanliness and the vice of drunkenness alike grow on what they feed upon. For in such cases our observations range over some period of time; and the man is not the same at the beginning as at the end of it." (Alfred Marshall, Principles of Economics, 8th ed. [London: Macmillan, 1962], 79.)

27. In his Ulysses and the Sirens (Cambridge: Cambridge University Press, 1979), Jon Elster's discussion of "The Locally Maximizing Machine" is similar to "myopic habit formation," and "The Globally Maximizing Machine" is akin to "purposeful preference molding." In his Sour Grapes (Cambridge: Cambridge University Press, 1983), Elster's "thin theory of individual rationality" considers only our "preference fulfillment effect" while his "broad theory of individual rationality" considers our "preference fulfillment effect" as well.

28. Von Weizsacker, "Endogenous Change of Tastes," 37 1.

29. Ibid.

30. Ibid.

3 1. Pollak, "Changes in Consumer Preferences," 374.

32. Pollak, "Habit Formation," 292.

33. The "impossibility" result derived independently by Pollak and ElSafty in the case of myopic habit formation stems from two considerations we will be careful to exclude in our treatment in chapter 6. (1) Habit formation makes nonconvexities highly likely. While true, we will waive the objection since it is well known that nonconvexities create problems for existence proofs. (2) Myopic adaptive behavior allows for the possibility that individuals will become "mired" in local optima that are not global optima. This is the principal focus of Pollak and El-Safty in their 1976 papers discussed previously. They demonstrate that a "local maximizing machine"-in Elster's words-will only succeed in global maximizing under highly restrictive assumptions about the nature of the local maximization process, namely the "separability" or "additivity" of the short-run utility functions. Since we will stipulate perfect foresight, we bypass this problem that has been the principal focus of the technical literature.

34. Pollak, "Habit Formation," 295.

35. Ibid., 294.

36. Ibid., 296.

37. Ibid.

38. We will leave to Pollak and others specializing in the field of empirical demand estimation to decide if myopic habit formation is more important, even if less tractable, to analyze than conscious preference molding when studying actual human behavior.

39. Thorstein Veblen and economists of the institutionalist school have long argued that actual consumer behavior can be better understood through the concepts of "invidious comparison ... .. conspicuous consumption," and "habit" or "inertia" than through the axioms of rationality. Let us assume for the moment that they are largely correct in this judgment. This would still not provide a cogent argument for abandoning an analysis of how an economy would function if people behaved in accordance with the axioms of rationality since this is the only way to determine the "welfare effects" of the economic institutions that "define" the economy. In a sense, Pollak is making the same error that many institutionalist economists have made in presuming a conflict of interest here.

40. Pollak, "Changes in Consumer Preferences," 377.

41. Ibid.

42. While Pollak is correct that unconditional preferences are the relevant preferences for welfare analysis and that such preferences cannot be inferred from market behavior-they are not the same as "revealed preferences'7--he is incorrect that little can be concluded regarding welfare in the context of endogenous preferences, as we will see.

43. Pollak, "Changes in Consumer Preferences," 374. We must point out that while Pollak admits to this problem in 1978, the insight was not his own. Gintis explained this very problem in just such language in numerous publications between 1969 and 1974 of which Pollak acknowledges being aware. But as we will see in chapter 6, both Pollak's and Gintis' intuitions were wrong in important respects regarding the relation between endogenous preferences and the fundamental theorems of welfare economics.

44. Ibid., 377.

45. Gintis is not the only author to conceptualize "purposeful preference molding" and grapple with the dilemmas they pose. Whereas GeorgescuRoegen and Strotz discussed the implications of endogenous preferences for "irreversible" demands in the 1950s, Harsanyi was the first, to our knowledge, to examine the welfare implications of what we call purposeful preference molding. Gintis' focus is very much in the tradition pioneered by Harsanyi in "Welfare Economics of Variable Tastes." More recently, Jon Elster has done much to bring the importance of this issue to the attention of a wider audience in a brief section on "Endogenous Change of Tastes" in chapter 2 of Logic and Society, in sections 5 and 6 of chapter 2 of Ulysses and the Sirens, in chapters I and 4 of Sour Grapes, and in his wn contribution in The Multiple Self, Jon Elster, ed. (Cambridge: Cambridge University Press, 1986).

We focus our attention here on Gintis because his work preceded Elster's, because our own intellectual debt is to Gintis rather than Elster whose work we only became aware of after formulating our own views on these matters, and because Gintis, as an economist, speaks the "language" of economists more consistently than Elster. But this choice is partly idiosyncratic on our part and we find ourselves very much in agreement with many of Elster's early views on these matters. We are less taken with the notion of "multiple selves," but this concept is admittedly aimed at a different set of problems than those we have found most intriguing.

46. Gintis, "Alienation and Power," 102-105.

47. Ibid., 105-106.

48. Ibid., 106.

49. Consistency requires that if bundle A is preferred to bundle B, then bundle B cannot be preferred to bundle A. This assumption is justified by traditional welfare theory by appealing to the intuitive notion of simple rationality. We agree that consistency is implied by rationality under the assumption of exogenous preferences. But under the assumption of en dogenous preferences lack of "consistency" may be perfectly rational. If bundle A consists of an Am radio and an adult education course in the appreciation of classical music, and bundle B contains an FM radio and an adult education course in the appreciation of classical music, I might very well "reveal" a preference for bundle A in the first time period and for bundle B in a later time period when the Am radio has worn out and the music appreciation course has had a chance to affect my auricular preferences.

Elster treats this issue in Ulysses and the Sirens, chapter 2, section 5 on "inconsistent time preferences." And possibilities such as this caught the attention of Georgescu-Roegen, "The Theory of Choice and the Constancy of Economic Laws," and Farrel, "Irreversible Demand Functions," Econometrica 20 no. 2 (April 1952): 171-86.

If it is admitted that this is perfectly reasonable or "rational" behavior that an adequate theory should be able to explain, the only way to salvage the axiom of consistency (much less transitivity) is to employ the artifice of telescoping a dynamic human world that moves unidirectionally through time into a static world of homo-economi making one single choice, in what could hardly be called their "lives," of all consumption and work bundles for every time period of their existence in the single choice period of the present. As we make clear later in this chapter, we agree with Gintis that to do this is to adopt a model that grossly violates important aspects of our actual human condition. On the other hand, we also believe that

adopting just such a model of endogenous preferences permits us to shed a great deal of light on the most important welfare theoretic questions concerning the important economic systems of our day. But all this is yet to come.

50. Gintis, "Alienation and Power," 118. In previous citations Pollak indicated that he also appreciated this complication.

5 1. Gintis, "Alienation and Power," 107-108.

52. As best we can ascertain, the first economist to elaborate the notion of utility depending on price was Thorstein Veblen in his famous notion of conspicuous consumption. To the extent people attempt to consume relatively more than others to achieve social status through "invidious comparison," traditional welfare theoretic conclusions are obviously misplaced. While the usual reaction has been to assume that goods consumed for status are relatively few and exceptions to the rule, some neoclassical economists have grappled with the complications they pose.

See Oscar Morgenstern, "Demand Theory Reconsidered," Quarterly Journal of Economics 62 (1948): 165-201; H. Leibenstein, "Bandwagon, Snob, and Veblen Effects in the Theory of Consumer Demand," Quarterly Journal of Economics 64 (1950): 183-207; R. Kalman, "A Theory of Consumer Behavior When Prices Enter the Utility Function," Econometrica 36 no. 3-4 (1968): 4975 10; H. Hayakawa and Y. Venerias, "Consumer Interdependence via Reference Groups," Journal of Political Economy 85 (1977): 565-99; and Robert Pollak, "Price Dependent Preferences," American Economic Review 67 no. 2 (March 1977): 64-75. Of course James Duesenberry's famous "relative income hypothesis" is rooted in a Veblenesque view of consumption. Income, Savings, and the Theory of Consumer Behavior (Cambridge, Mass.: Harvard University Press, 1949).

53. Gintis, "Alienation and Power," 106.

54. Ibid., 106-107.

55. Although the reason is ultimately of no consequence, perhaps Gintis was overly influenced by his study of Mill, who placed great emphasis on the problem of uncertainty in the analysis of preference development.

56. Gintis, "Alienation and Power," 99.

57. Ibid., 127.

58. Ibid., 100.

5 9. Ibid., I 10.

60. Ibid., 112.

6 1. Ibid., 113.

62. Ibid., 119, 121, 122, 127.

63. Ibid., 127.

64. Ibid., 262.

65. Gintis, "Welfare Criteria with Endogenous Preferences: The Economics of Education," International Economic Review 15, no. 2 (June 1974): 426, Theorem 5.

66. Ibid.

67. Ibid., 427.

68. Ibid., 426.

69. See our treatment of these issues in chapter 9 on central planning.

70. Gintis, "Alienation and Power," 18.

71. Roy Harrod, "Scope and Method of Economics," Economic Journal 48, no. 191 (September 1938): 386.

72. Joseph Schumpeter is perhaps the best-known proponent of this view.

73. This contradiction has been treated in a subfield of traditional economic theory regarding patents and copyrights and is not really resolved. Rather, the consequences and best practical course of action are debated. We suspect, however, that the newest technological revolution in microcomputers and software will focus much more attention on the issue. When the "commodity" is knowledge in the form of a program recorded on a diskette that can be almost costlessly transferred from the first buyer to all potential subsequent buyers; when the market is "mass" rather than a small number of business and institutional clients whose subsequent transfers can be reasonably policed; and as long as the U.S. Constitution makes searches of tens of millions of homes legally difficult as well as economically costly; we have the makings of an interesting test of how optimal the effect of private enterprise market institutions on technological progress really is. We hasten to point out that it should not be assumed that the outcome would be the same in a different institutional setting.

It also bears mentioning that French, German, and Japanese "visions" of the role of the state in fomenting technological change and structural transformations in a private enterprise market economy has always differed markedly from Anglo-American laissez-faire conceptions. Obviously, some interesting "success" stories are on the opposite side of the great ideological divide. See Andrew Shonfield, Modern Capitalism (London: Oxford University Press, 1974) for an interesting, though dated, comparison of the French and English experiences.

74. See Gintis' discussion of Marshall in "Alienation and Power," 40-41.

75. John Sawyer, "Social Structure and Economic Progress: General Propositions and Some French Examples," American Economic Review 41 no. 2 (May 1951): 321-29.

76. Ibid., 321-22.

77. Ibid., 322.

78. Ibid., 323.

79. Ibid., 324.

80. Ibid., 322-23.

8 1. Of course, to the extent people do not follow the dictates of individual rationality, the social impact of institutions will be less. But the critical word here is "less," The impact will be "less" to the extent actual behavior deviates from "promoted" behavior, but not "different." In the extreme, one might argue that if there were literally no correlation between rational and actual behavior there would be no "scientific" purpose to welfare theory. But in this case there would be no "scientific" purpose in studying any social institutions, or anything else in the "social sciences" for that matter.

82. Oscar Lange and Frederick Taylor, On the Economic Theory of Socialism (New York: Monthly Review Press, 1964). An interesting feature of the LangeLerner-Taylor model is that optimality properties do not depend on market structures being atomistic or competitive. Their use of a Central Planning Board for making price adjustments in intermediate markets is basically irrelevant from a welfare theoretic point of view if we are only interested in equilibria in any case. Whether their adjustment mechanism is an improvement on the "natural" market mechanism might, however, be of interest to modern, post-Keynesian disequilibriurn theory.

83. Traditional theory equates "socialism" with "public enterprise economies." While at one level it is a matter of semantics, we prefer to define "socialism" to mean an economy in which those who carry out economic activities are in a position to conceive and coordinate those activities themselves. But if we use the term this way, neither CPS nor WSMS can be deemed socialist (see Michael Albert and Robin Hahnel, Socialism Today and Tomorrow [Boston: South End Press, 1981)). The issue transcends semantics because by attaching the label "socialist" to countries such as the Soviet Union, China, etc., the rulers of those countries attempt to legitimate actions that pursue wealth and power for an elite as "serving the people," and denigrate criticism as "attacks on working people." Attaching the label "socialist" to these countries also permits rulers of capitalist countries to denigrate the label and create a context in which they claim all who criticize capitalism are pro-Gulag. In short, equating "socialism" with existing public enterprise economies is part of the Orwellian process of constricting what can be thought in a world where thoughtwithout-limits would, as always, be a threat to the powers that be.

84. To avoid misinterpretation, when we say all the above systems are equally efficient, we mean each is capable of generating Pareto-optimal outcomes under appropriate assumptions. In general, these assumptions concern perfect information for decision makers, faithful execution of all decisions, and, in market systems, ability to achieve a general equilibria. But the decision makers are different in the different systems, and the information they are likely to possess and the incentives they face may be different as well. Therefore, formulation of the assumptions necessary for different systems to achieve Pareto optimality differ. All of which has been the subject of what we call the "modern debate" over the information and incentives properties of economies. But while we recognize that "practical" considerations such as these certainly merit discussion, we find the necessary assumptions for the different systems essentially equivalent.


1. An interested reader might well begin with Michael Albert, Leslie Cagan, Noam Chomsky, Robin Hahnel Mel King, Lydia Sargent, and Holly Sklar, Liberating Theory (Boston: South End Press, 1986).

2. In our opinion, paradigms in the social sciences should not be judged entirely on the basis of immediate results. Because so many unanswered questions always remain, using this criteria alone is too risky. While a new paradigm may prove fruitful in some areas, internal weaknesses may present roadblocks to progress in other areas later. This is precisely what we claim has happened with the traditional welfare paradigm. So it is important to continuously requestion our underlying assumptions about human beings and society in light of new scientific knowledge-a process that is always controversial and subject to ideological influences. However, in a book proposing a new approach it seems sufficient to present a new paradigm that yields interesting new results-the most important first step in any case. Those who are sufficiently interested will no doubt have to examine the merits of the new paradigm more closely to weed out inaccuracies and add new concepts if the paradigm is to continue to facilitate progress.

3. See Peter Kropotkin, Mutual Aid: A Factor of Evolution (New York: New York University Press, 1972).

4. If human society and our environment remained very different from neolithic conditions and persisted for an equally long period, we would have to consider further evolution of "human nature." But more pressing matters are likely to change things first. If increasing understanding of the evolutionary process does not change the nature of our subjection to its "laws," advances in genetic engineering certainly will. But these issues are not important to our present purposes.

5. Our view of human nature owes immeasurably to the work of Noam Chomsky. Readers might consult any of his more popular books on linguistics, for example, Language and Problems of Knowledge, The Managua Lectures (Boston: NUT Press, 1988).

6. We wish to practice no sleight of hand here. The fact that we do not have wings distinguishes us from pigeons. The fact that we can mate only with other humans distinguishes us from all other species. But these "physical" genetic differences are not what is critical to a human nature-based approach to welfare theory. The key assertion is that part of human nature includes particular intellectual, psychological, 'emotional attributes, and potentials that have important implications for what social arrangements will prove "stable" and "fulfilling." This is the approach we develop.

7. See David Cohen, ed., Psychologists on Psychology, (New York: Tapplinger, 1977) for a critique of Skinnerian approaches to psychology and policy-making.

8. Richard Lichtman, "Marx and Freud: Marx's Theory of Human Nature," part 3, Socialist Revolution 7, no. 36 (November-December 1977): 45. To avoid any misrepresentation, we should note that Lichtman argues against this interpretation of Marx's views on the subject. However, he argues that this is, in fact, the interpretation of most Marxists.

9. Karl Marx, Capital, Vol. 1, The Process of Capitalist Production (New York: International Publishers, 1967), 178.

10. Even if, as Kelvin Lancaster suggested, one defines utilities as functions of activities rather than commodities, or defines preference orderings on an activity space rather than a commodity space, this problem persists. The question of how a particular activity was selected and how society determined that access to different activities and their consequent enjoyment was to be distributed is of welfare significance for humans.

11. It is worth noting that Jurgen Habermas and many of his disciples consider communicative structures as important as decision-making structures. When they assess the processes as well as results of social decision making, they find the fairness and efficiency of communication systems as important as the rules of decision making. The relevant questions are: Who has access to what information? What means for conveying information and opinions exist? How is the communicative interchange organized? These are some of the interesting questions the "modern" Frankfurt school has focused on. Remarkably, just by setting minimal goals for "humane communication," members of this school have elaborated guiding values for much of society.

12. Peter Marin, "The Human Harvest," Mother Jones (December 1976): 38.

13. Bertell Ollman, Alienation (Cambridge: Cambridge University Press, 1973),108.

14. Karl Marx, Introduction to a Critique of Political Economy (New York: International Publishers, 1972), 268.

15. Marin, "Human Harvest," 32.

16. As we saw in chapter 4, the unconscious way in which present actions can influence future preferences was the primary concern and focus of economists such as Pollak who focused on myopic habit formation.

17. While traditional welfare theorists have not focused attention on this dynamic, it is not difficult to extend traditional theory in ways that do so. As we will see in chapter 6, once preferences are treated as endogenous, the same analysis applied by traditional welfare theorists in human capital theory facilitates analysis of this trade-off. Moreover, the so-called "wealth effect" used to explain why those with limited means may rationally not insist on as great "compensating wage differentials" for undesirable job characteristics as those who are better-off applies here as well. See Norman Waitzman, "The Occupational Determinants of Health: A Labor Market Segmentation Analysis" (Ph.D. diss., American University, 1988) for an excellent evaluation of this approach to health risks.

18. The rather surprising extent to which this argument can be developed without making specific assumptions concerning the nature of people's preferences is a primary concern of chapter 6. But suffice it to say that the entire argument is totally foreign to traditional welfare theory and beyond the vision of the traditional paradigm.

19. Individuals choose from a limited menu of activities. Choice of menus is a social choice of how social activities will be organized and packaged by major social institutions.

20. It is sufficient for now to indicate how different welfare issues will look from the perspective of our new paradigm. Justifying the merits of our new approach is the subject of the rest of the book.

21. See Joseph Weizenbaum, Computer Power and Human Reason (New York: Freeman and Company, 1976); and David Noble, America By Design (New York: Knopf, 1977). We emphasize institutions and roles in our discussion of societal boundaries, but artifacts can have considerable importance in their own right. In some circumstances, the "boundary" of "artifacts" is so complex to use or replace or imposes such powerful operational dictates that the objects themselves delimit role possibilities.

22. This point is relevant to some naive interpretations of anarchism. Human social life without institutions is impossible. Consequently, the vision of human liberation as the abolition of social institutions is a deceptive pipe dream. The intuition common to anarchist and existentialist thought that all social institutions mold, constrain, and limit individual human choice, no matter how gently and subtly, is well founded. But the conclusion that liberation means abolition of these constraints on individual action does not follow once we recognize that this is humanly impossible. See Albert et al., Liberating Theory for further discussion of a more useful conception of "liberation" as the informed and democratic choice of social institutions that permit maximum development and fulfillment of human potentials.

23. In our view, Thorstein Veblen, founder of the institutional school of economics, and Talcott Parsons, a giant of modern sociology, both underestimated the potential for applying the human tool of consciousness to the task of analyzing and evaluating the effects of social institutions. This led Veblen to overstate his case against what he termed "teleological" theories of history and rendered Parsonian sociology powerless to explain the process of social change. While many could be cited on this subject, we find the work of Anthony Giddens particularly useful in avoiding the pitfalls of unjustifiable "functionalism" and sterile behavioral sociology. We hope we have applied Giddens' insights in constructing our paradigm and theory.

See Anthony Giddens, Central Problems in Social Theory: Action, Structure, and Contradiction (Berkeley: University of California Press, 1979); Contemporary Critique of Historical Materialism (Berkeley: University of California Press, 1981); Social Theory and Modern Sociology (Cambridge: Polity in association with Blackwell, 1987); and Social Theory Today (Cambridge: Polity in association with Blackwell, 1987).

24. See Albert et al., Liberating Theory.

25. The distinction between primary and produced inputs is also critical in studying the relation between price determination and income distribution in private enterprise market economics. For an excellent treatment see G. Abraham Frois and E. Berrebi, Theory of Value, Prices, and Accumulation (Cambridge: Cambridge University Press, 1979).

26. Of course, human economic efforts are often required to discover and make available natural resources we define as "nonproducible." And in some cases human processes can produce facsimiles of nature's products. With sufficient crushing, shale will become oil, and coal will become diamond. And sometimes it is possible to develop producible "synthetic" substitutes for dwindling resources. In any case, we do not suggest the distinction between produced inputs and resources to ignore these possibilities, but simply to allow the framework to assess properly the impact of economic choices that deplete resources or otherwise alter ecological balances.

27. While it is often not important for welfare theoretic purposes to account for the duration and intensity of efforts carried out in consumption activities, we discuss important exceptions below.

28. There are additional conceptual and practical problems in defining and measuring "services" as outputs, but we have no particular contributions to make on this subject.

29. The levels of fulfillment, which we define as the human outputs of economic activity, are not the only human consequences of economic activity. Indeed a critical difference between an endogenous as opposed to exogenous conception of preferences is the potential transformation (or reinforcement) of human characteristics that results from engaging in an economic activity. But rather than call the transformations of human characteristics human outputs, we will analyze these highly important human consequences as changes in the values of human state variables.

30. "On the job training" is one of the ways this far more general phenomenon has received attention. A transformation deemed to be an improvement in workers' specifically job related skills is focused on as one of the results of organizing and carrying out the activity in a particular manner.

3 1. We do not mean to imply that all activity of women in the home should be conceptualized through economic metaphors. Quite the contrary, we are highly critical of "economistic" interpretations that presume these activities will only be taken seriously if interpreted as "economic production." Elsewhere we have argued strenuously that many of these activities are best conceptualized as kinship activity-activity oriented primarily toward satisfying sexual, effective, and child-rearing needs organized according to a particular (usually sexist) sex/gender system or division of responsibilities and rewards-and that they are no less important for being so.

But some activities usually assigned to housewives are appropriately conceived as primarily economic. We submit that in most such cases they fit the category of "consumption activity" better than "production," and that our paradigm provides for a serious treatment of those efforts just as it does "productive" efforts. To put it differently, the traditional paradigm, and some of its better-known competitors generate a biased attitude toward economic as opposed to noneconomic activity, and toward production as opposed to consumption activity. Only if one denigrates the importance of kinship and consumption activity is there a problem with applying these labels to activity in the home primarily carried out by women.

The paradigm presented in this chapter attempts to overcome both these biases. In our view many treatments of "housework" by Marxist and Marxist/Feminist authors in the 1970s incorporated the biases and resulted in skewed analyses. For one example by an author with whom we are in agreement on most matters, see Ann Ferguson, "Women as a New Revolutionary Class," in Between Labor and Capital, Pat Walker, ed. (Boston: South End Press, 1979). In Michael Albert and Robin Hahnel, Marxism and Socialist Theory (Boston: South End Press, 1981), we elaborate our criticism of this literature.

32. In this regard economists have adopted the approach of philosophers rather than psychologists. For most of us this is an unconscious choice. In our view it is far more justifiable for welfare theory than the so-called "analytical" theory of consumer behavior. But in any case, it is interesting to note that the "basic needs" school of development theory is perhaps the only area in which economists adopt a psychological paradigm rather than a philosophical one. Abraham Maslow's hierarchy of needs is an admirable paradigm for economists working in this genre.

33. Readers might inquire about the relationship between our concept of an "activity possibility set" and the traditional concept of a firm's "production possibility set." The traditional concept is aptly described as a book of blueprints in which each page represents a way to turn a vector of inputs into a vector of outputs known to a particular firm. Since in the traditional conception no particular people inhabit firms, it is a little difficult to know "to whom" these technologies must be known, or why the book of blueprints would not be the same for all firms. In any case, in its usual variant, the production possibility set is limited only by technologies known to a firm at a particular time. In contrast, A(i,t) is limited not only by technologies known to the members of a particular group i, at time t, but by individual and group human characteristics of group i at time t. For example, suppose a particular activity of which group i was fully aware required behavior members of group i were either individually or collectively incapable of carrying out. The activity could not be undertaken.


1. Truly individual activities are incorporated in this formalization as activities carried out by a group, i, with a single member.

2. Differences in intensity can be handled in either of two ways. Hours of indoor carpentry at intensity A and hours of indoor carpentry at intensity B can be defined as two different elements in "work space." Alternatively, we can define hours of indoor carpentry as a single element of work space and attach the appropriate intensity rating to that element before applying our instantaneous preference ordering to evaluate the welfare effects of the work activity, as was explained in chapter 5.

3. Insisting that expenditures be covered out of current income introduces the complication that an individual's marginal utility of income will probably differ in different time periods. While this is somewhat bother some, the additional complexities of permitting individuals to borrow and lend are more burdensome still, and in no way affect any of our results.

4. Readers familiar with the neoclassical literature on endogenous preferences will no doubt be struck by the difference between the conclusions we derive and the conclusions published by Pollak, El-Safty, Hammond, and Gintis. In most cases this is not because their conclusions do not follow from their premises, as we hope is the case for us as well. Pollak, El-Safty, and Hammond found that only in very restricted cases would individual optimization yield anything we should consider a welfare optimum, because they assumed myopic adaptive behavior in which individuals optimize at each moment according to their instantaneous preference ordering ignoring the preference development effects of these choices. Moreover, by concentrating on habit as the kind of endogeneity, they (correctly in their case) envisioned serious existence problems due to nonconvexities. Gintis did err in concluding that many of the multiple equilibria he envisioned would be nonoptimal because, while he conceded a perfect knowledge counterpart to the CPS, he lapsed into thinking in the same terms as Pollak, EI-Safty, and Hammond-namely of myopic, local, maximizing machines. Our conclusions follow rather directly from our model of endogenous preferences with perfect knowledge since these effectively create global maximizing machines insightful enough to scan all individual local maxima and select only individually global maxima.

5. This proposition was first demonstrated by Herb Gintis, "Welfare Criteria with Endogenous Preferences: The Economics of Education," International Economics Review 15, no. 2 (June 1974): 422, Theorem 3.

6. If some of the budget constraints were nonbinding, or if the solution point were not interior, we could model the individual's maximization problem as a general nonlinear programming problem and analyze the full set of KuhnTucker conditions. However, since nothing relevant to the discussion is lost by considering the simpler case, for convenience we will do so.

7. We ignore the complication of discrepancies between private and social cost and individual and social benefit for now.

8. While this assumption may appear more restrictive than modern traditional treatments, we do not believe this objection is consequential. We have chosen to present this part of our argument using the admittedly outdated concepts of utility functions and marginal utilities simply because the exposition is easier. But none of our results depend on this simplification. In this context, our assumption of diminishing marginal utilities is merely the analogue of the assumptions of diminishing rates of substitution in consumption or convex sets of at-least-as-preferred-as consumption bundles-common assumptions in modern traditional treatments necessary for deriving well-known results.

9. K. Arrow and F. Hahn, General Competitive Analysis (San Francisco, Holden-Day, 1971), chap. 6, briefly treats the case where prices affect utilities in the sense of conspicuous consumption. But the sense in which utility depends on price due to our conceptualization of endogenous preferences is entirely different from the cases analyzed in Arrow and Hahn, and their analysis is, therefore, not relevant to our concerns.

10. Quite to the contrary, we will argue in chapters 7 and 8 that there are major problems with both private enterprise and market institutions, which in combination with endogenous preferences produce a snowballing nonoptimality of double proportions in even the most perfectly competitive, private enterprise market economies.

11. However, as we explain below, there is good reason to believe that extending the convexity assumptions to endogenous well-being functions stretches credulity more than under traditional conditions. Put differently, it appears additional reasons to fear nonconvexities when preferences are recognized to be endogenous make the usual convexity assumptions less plausible.

12. Gerard Debreu, Theory of Value (New York: Wiley, 1959), chap. 5; or K. Arrow, "General Economic Equilibrium: Purpose, Analytic Techniques, Collective Choice," American Economic Review 64, no. 3 (June 1974): 253-72.

13. See Debreu, Theory of Value, chap. 4, sec. 4.8 for treatment of these technical complications.

14. Ibid., sec. 4.9.

15. Arrow and Hahn, General Competitive Analysis, chap. 7.

16. Peter Dorman is a notable exception who has developed this theme recently in a number of working papers soon to be published.

17. See Emile Durkheim, The Rules of Sociological Method, 8th ed. (Chicago: University of Chicago Press, 1938); and Max Weber, The Theory of Social and Economic Organization (New York: Oxford University Press, 1947) for compelling arguments from two founders of modern interactive social theory .

18. Arrow and Hahn, General Competitive Analysis, chap. 7; and R. N. Anderson, M. A. Khan, and S. Rashid, "Approximate Equilibrium with Bounds Independent of Preferences," Review of Economic Studies 44 (1982): 473-75.

19. Debreu, Theory of Value, sec. 6.3. 20. Ibid., sec. 6.4; and Arrow, "General Economic Equilibrium," 263-64.

2 1. Readers still plagued by visions of "multiple self-fulfilling equilibria," any of which are socially inefficient, should recall the assumptions

inherent in Debreuvian models of perfect foresight and a single choice for all time periods, which was the model Gintis finally agreed to use when drawing welfare conclusions, i.e., his "complete knowledge counterpart to the individual's CPO." Gintis' socially inefficient nightmares are effectively ruled out by individuals who would reject them as individually nonoptimal when scanning all possible trajectories prior to making their choices for each and every time period before beginning. Pollak, EI-Safty, and Hammond's conclusions that "long-run demand functions can only be rationalized by a utility function if and only if the short-run utility function is such that any good that experiences learning or taste change is separable from all other goods" (A. EI-Safty, "Adaptive Behavior, Demand and Preferences," Journal of Economic Theory, no. 13 [1976]: 298) also stems from their decision not to grant individuals perfect knowledge of their "meta" utility function so they can maximize using this information, but instead must maximize "myopically" with their short-run utility functions incognizant of preference development effects.

22. When we say the economy was a marshmallow influenced only by people's "economic wills" we do not mean it could furnish any technology, supply of natural resources, or initial human characteristics people might want. We mean with what traditional economic theory calls a particular set of "givens," the economy was a marshmallow to be molded entirely by people's wills.

23. It should be clear that endogenous preferences have nothing to do with the economic system. They are a property that people can be presumed to have or not have. If private enterprise market economies generated no inefficiencies, recognizing that their inhabitants are imbued with this capacity could not make them do so!

24. We also assume no exchange of goods between firms and consumers can increase a consumer's satisfaction without decreasing a firm's output or increase a firm's output without decreasing some consumer's satisfaction. In other words, we assume no misallocation of goods between use in production and consumption.

25. Clarification is in order because "net marginal social costs" would be redundant in some conceptualizations. We define the "marginal social cost" of supplying an activity to be all additional costs to society of making an additional unit available irrespective of who consumes it. These costs would include not only the increase in costs to the supplier, but any additional costs to society that the economic system, for whatever reason, fails to charge the supplier, e.g., the costs of pollution as well as inputs that must be paid for. Marginal social costs and marginal private costs would be equal if and only if there were no externalities in production activities. We define "net marginal social cost" as the marginal social cost minus the

benefits all others enjoy as a result of the fact that the good has been made available to a particular individual. In "private" activities the two are the same. But in "consumption" activities with "external" or "public" effects, marginal social costs and net marginal social costs will differ. In other words, net marginal social costs and marginal social costs would be equal if and only if there were no externalities in consumption activities. So while marginal social cost is determined entirely by conditions within the production sector, net marginal social cost depends on the nature of consumption activity and consumers' preferences as well. Moreover, the net marginal social cost of something will, in general, vary depending on who "consumes" it. The net marginal social cost of supplying person i with another unit of activity b will be different from the net marginal social cost of supplying person j with another unit of activity b if i and j do not benefit equally from the other person's consumption of b.

26. Note that definitions (M 3.3) could have been stated in terms of enhancing appreciation of one activity relative to the other and developing one characteristic relative to the other since it is perfectly possible an action might have developmental or appreciation consequences on more than one characteristic or good. In this case (M 3.3) would specify: engaging in activity y increases characteristic S more than P as compared to engaging in activity x; engaging in activity x increases characteristic P more than S as compared to engaging in activity y; characteristic S enhances appreciation of activity a relative to activity z as compared to characteristic P; and characteristic P enhances appreciation of activity z relative to activity a as compared to activity S. But formulating the model in this way only complicates matters without enriching results.

27. This is a further simplification. In our model of an individual with endogenous preferences, we assumed the overall individual welfare function was a linear weighted sum of utility in each time period where the weights represented the individual's subjective rates of time discount. Here we specify weights of one, thereby assuming a zero rate of time discount. While such an assumption might not be farfetched for the immortals we have created, the real point is that none of our results are affected by the additional simplification.

28. Of course if providing i with a unit of b or c to consume does not affect the utilities of others, the ratio of net marginal social costs reduces to the ratio of marginal social costs. Some might object that in this formulation of net marginal social costs we have only included the immediate preference fulfillment effects on well-being forj # i, and not incorporated the future preference development effects on well-being. While true, the additional complication does not affect our results. Moreover, since our proof considers the production of goods a and z, neither of which have developmental effects, the formulation is accurate for the case we will treat.

29. If, and only if, all individuals are charged the net marginal social costs of providing them with activities will the sum of individual demands that yields the market demand be such that marginal social costs will be equal to marginal social benefits. To see this, if i is charged an amount equal to the net marginal social cost of providing him or her another unit of an activity, under competitive conditions i will demand that activity up to the point where the marginal private benefit is equal to the net marginal social benefit thereby establishing: { Marginal Benefit to i Marginal Social Cost Marginal Benefit to all others). But in this case: Marginal Social Cost Marginal Benefit to i + Marginal Benefit to all others) = (Marginal Benefit to all), or Marginal Social Benefit. And clearly, charging i anything different than the net marginal social cost of the activity will leave Marginal Social Cost not equal to Marginal Social Benefit.

30. We make the additional assumption of no comer solutions here to simplify the argument, but note that all our conclusions are valid without this assumption as well.

31. As explained above, the "optimal" prices to charge individual purchasers are the net marginal social cost of supplying the individual with another unit of the activity.

32. When we say neoclassical theory here, we mean traditional welfare theory, which treats preferences as if they were exogenous.

33. The logic of this adjustment is, of course, a primary subject of traditional price theory, and the conclusion is warranted except in the case of "Giffen" goods.

34. In part we did not assume this because it was not necessary for our conclusions. In part we did not assume this in order to distinguish our analysis from any kind of habit formation, be it myopic or not.

35. Excluding technical concerns, the assumption of endogenous preferences does not affect theorems guaranteeing the existence of general equilibria in market systems, or the ability of a procedure to generate feasible plans in planning systems. We proved this result for market systems as Theorem 6.3 and will prove it for centrally planned systems in chapter 9. We relegate this part of our summary to a footnote because when we speak of "any economy," as we did previously in the text, this presupposes a system that has a general equilibrium or can produce a feasible plan. Any model we theorized that did not have a general equilibrium or generate a feasible plan could not "be" an economy and should be discarded as an internally inconsistent model that cannot represent "an economy"even in theory.

36. Rawls stated this qualification as follows: "Each person is to have an equal right to the most extensive basic liberty compatible with a similar liberty for others." (John Rawls, A Theory of Justice, [Cambridge, Mass.: Harvard University Press, 1971], 60.)

37. We do not mean to imply that this issue could be easily resolved by empirical studies. In our view paradigms do have a legitimate role to play in just such situations.

38. In a slightly different conception-more applicable to artisans than farmers-each actor is seen as being capable of producing the tools required before going on to use them to produce whatever products he or she will eventually consume or trade. This reduces the concept of "wherewithal" entirely to personal capacities since now no specific tools are required to engage in an activity; they can be self-produced prior to the activity-or essentially as part of a larger redefined activity. Once this is accomplished, all that is necessary to permit individuals to switch from one "choice activity set" to another should they so desire is to grant all the capability to learn how to make the tools required to produce any product, In other words, in its most operational form the classical liberal paradigm renders individual economic choice activity sets not only independent but identical. That neoclassical theory tells us the outcome of an exchange economy with identical production possibility sets is indeterminate has seldom been seen as contradictory.

39. In terms of our qualitative model, the more "input flows" an activity requires to be supplied from other activities, the less "self-sufficient" it is. Or, put differently, the greater "division of labor" it presumes.

40. B is in no position to infringe upon the liberty of A if we assume that B cannot make A do what would enable B to do what B would like to do as B's part of the division of labor. Of course, if B can compel A to carry out A's part in the division of labor that B prefers to mutual autarchy, then B is infringing on A's liberty in an obvious way. We wished to point out the less obvious way in which A infringes upon B's liberty by choosing to act "independently."

41. This is commonly referred to as "left adventurism."

42. This "error" is commonly referred to as "right opportunism."

43. At the risk of sounding ridiculous to those outside the profession, an economists' explanation of the well-being generated by closely knit families would be that family solidarity creates powerful positive externalities that multiply the overall well-being that results from the good fortunes of individual members-and we are not referring to prolonged visits of poor relations and zero interest rate loans between kin.

44. See the chapters on "community" in Michael Albert and Robin Hahnel, Marxism and Socialist Theory (Boston: South End Press, 1981); and Michael Albert et al., Liberating Theory (Boston: South End Press, 1986).

45. See our treatment in chapter 1.

46. It is interesting to note that according to Rawls any objection to an increase in inequality that improved the position of the worst-off could only be the result of envy. On these grounds Rawls refuses to sustain any objections to increases in inequality that have the effect of improving the "objective" position of the worst-off. Our inclination is to be unable to accept great deviations from material equality as anything other than evidence of failure to apply equality of consideration sincerely. After all, "super people"somehow deserving of much greater material reward for their participation in joint social endeavors--only exist in the context of a particular set of social conditions, values, and reward structures. Ultimately the development of solidarity requires eliminating any circumstances that translate differences in talents or skills, that most certainly do, will, and should exist, into differences in material well-being and power. In other words, there would be Magic Johnsons and Albert Einsteins whose talents are appreciated. But they would not be paid millions. In our view, since failure to rectify the circumstances that result in inequalities--even should they improve the material well-being of the worst-off-"objectively" limits solidarity, we have an "efficiency" reason for possibly sustaining objections to such inequalities. But perhaps the difference is only one of semantics: what we see as "refusal of the disadvantaged to have solidarity with those who don't put their money where their mouth is," Rawls calls "envy."

47. Robert Nozick, Anarchy, State, and Utopia (New York: Basic Books, 1974),310.

48. One of the authors is a parent of identical twins. Besides sharing the same genes they have been each other's most important "environment" both inside the womb and since birth as well. They could have developed similar characteristics, but instead they have developed quite differently (which should come as no surprise to anyone who has known genetically identical siblings). The claim here is that by developing very different characteristics, and thereby providing one another with very different "environments," they have "optimized." Not that things have turned out perfectly! But identical developmental trajectories would have been a nightmare. More seriously, the psychological literature on twins affirms the importance of different development trajectories.

49. One does not have to assume "risk aversion" on people's part to deduce that it is rational to minimize risk in this way!

50. For example, Veblen clearly saw an inherent contradiction between what he called the human instinct to workmanship and the acquisitive instinct. In our human nature-based approach we postulated no such "acquisitive instinct," nor did we postulate any "aggressive" instinct or

innate need to dominate. We find institutional explanations more compelling explanations for the ubiquity of acquisitive, aggressive, power-seeking behavior and think these kinds of behavior are best explained as the outgrowth of desires to satisfy "natural" needs for security, social esteem, and self-management under particular "contradictory" social conditions. But others may disagree.

51. By "antisocial behavior" we mean behavior that is contrary to the interests of species survival, not behavior in opposition to established social norms and institutions which, depending on the "social rationality" of the norms, may or may not be contrary to the interests of species survival.

52. Evidence in favor of this view will be presented in part 3.

53. When we say "believe" we understand that belief may be unconscious or implicit. Indeed one of our major contentions is that very often the most important "beliefs" of theorists are unconscious and implicit.

54. By our treatments of classical utilitarianism, modern neoclassical theory, and modern contractarianism in chapter 1, it must have been obvious we recognized differences between them. As we were careful to explain, the entire philosophical basis of modern contractarianism is different from utilitarian and neoclassical theory. Why then did we lump them together under the label "traditional welfare theory"? The full reason can only now be apparent. What we have just explained is that their implicit conception of preference formation is the same. In other words, in this respect they are based on the same paradigm-the traditional paradigm we have criticized and proposed replacing. We believe that those who have used modern contractarian concepts have also accepted other aspects of the traditional paradigm such as the presumption that private effects are pervasive and external effects exceptional. Classification of modern contractarianism as part of traditional welfare theory does not rest on its treatment of preferences (or "individual's conceptions of the good") as exogenous alone. The criterion we used for classification purposes was the paradigm accepted by those using the approach, which they sought to express in different theoretical form. The philosophical form the theory takes was not our criterion for classification purposes, although it might well be for other purposes.

55. If the joys of conspicuous wealth, aggression, and domination are so attractive that people are loath to choose paths that do not afford them even small probabilities of enjoying these joys, why not consider "slave for a day" type solutions? Take the presumed aggregate benefits of enhancing dominating tendencies and instead of permitting people to "win" or "lose" individual benefits permanently, we could insist on temporary rich (and poor), temporary mugger (and muggee), and temporary boss (and worker) outcomes. It is curious this has seldom occurred to those who claim to

prefer a rugged, roll-the-dice, Horatio Alger, John Wayne world. But then many argue for this from the heights of power and with the benefit of hindsight, rather than from the original position behind the veil of ignorance.

56. There is a difference between the debate over the maximin principle versus maximizing expected benefits and what we consider here worth noting. The maximin principle justifies any change that improves the position of the worstoff. Maximizing expected benefits justifies any change that generates positive aggregate benefits. The maximin principle could fail to approve a change that had positive aggregate benefits, just as maximizing expected benefits could fail to approve a change that would improve the position of the worst-off. But the cases we discuss here all generate aggregate benefits by assumption. What we are comparing are on the one hand development trajectories that not only generate aggregate benefits (and, therefore, increase expected individual benefits) but are Pareto improvements as well (greater self-management and solidarity), as compared to trajectories that increase expected individual benefits (by assumption) but are not Pareto improvements because some will become worse off (greater acquisitiveness and aggression). It is still possible that people who are not "risk-averse" would choose the latter over the former, but in our view this "counterargument" is weaker in our context than in the well-known debate over Rawls' maximin principle.

57. See Jon Wisman, "Economic Reform for Humanity's Greatest Struggle" (the prize-winning essay tee the $10,000,1987 Speiser Essay Contest), for a highly persuasive argument for self-management on precisely these grounds. A slightly adapted version was published as Jon Wisman, "An Economic Response to the Threat of Nuclear War" Forum for Social Economics, Villanova University College of Commerce and Finance and Temple University School of Business and Management, 1987.


1. Ernest Mandel, "In Defense of Socialist Planning," New Left Review, no. 159 (September-October 1986): 5-38; and Alec Nove, The Economics of Feasible Socialism (London: George Allen and Unwin, 1983). Lest there be any doubt about which way the wind blows still, see Diane Elson, "Socialization of the Market," New Left Review, no. 172 (NovemberDecember 1988): 3-44.

2. See C. J. Bliss, "Prices, Markets, and Planning," Economic Journal 82 (March 1972): 87-100, for an excellent example of the terms of the debate from the perspective of a traditional critic of markets. See Allan Buchanan, Ethics, Efficiency, and the Market (Totowa, N.J.: Roman and Littlefield,

1985), chap. 2, for an excellent summary of "Efficiency Arguments For and Against the Market," 14-46, from the traditional perspective. Charles E. Lindblom, Politics and Markets (New York: Basic Books, 1977) is another well-known treatment.

3. In this case, as in many others, "new" ideas are seldom completely new. The market perfection theorem was very much a theme of the Austrian School, which had long objected to the traditional neoclassical truce regarding markets. See Don Lavoie, "A Critique of the Standard Account of the Socialist Calculation Debate," The Journal of Libertarian Studies 5, no. 1 (Winter 198 1); and idem, Rivalry and Central Planning: The Socialist Calculation Debate Reconsidered (Cambridge: Cambridge University Press, 1985) for a modern discussion of themes emphasized by von Mises and Hayek.

4. Oscar Lange and Frederick Taylor, On the Economic Theory of Socialism (New York: Monthly Review Press, 1964).

5. To contrast PuEEMMEs with more familiar private enterprise market systems: in PrEMEs buying shares of stock gains entrance to stockholders' meetings where one can cast as many votes as one's shares of stock on any proposition. But in employee-managed enterprises, simply being an employee gives you the same number of votes in the workers' council as every other employee-and there are no other voters.

6. The same critics also claim that equilibria of employee-managed market economies are less stable than equilibria of their private enterprise counterparts and that adjustment dynamics are perverse. In our opinion, this conclusion is also more a product of inappropriate choice of model than of any actual tendencies of PuEEMMEs. But we do not treat this issue here. In models where labor is the only input in production, and no entry and exit of new firms is permitted, adjustment in PuEEMMEs is certainly more problematical than in PrEMEs. But we have chosen to ignore all problems of dynamic adjustment of market systems in an effort to give market economies every benefit of the doubt and focus on other concerns. In this spirit we ignore a major component of the literature critical of PuEEMMEs.

7. Benjamin Ward, "The Firm in Illyria, Market Syndicalism," American Economic Review 48 no. 4 (September 1958): 566-89; idem, The Socialist Economy: A Study of Organizational Alternatives (New York: Random House, 1967), chap. 8; Evsey Domar, "The Socialist Collective Farm as a Producer Cooperative," American Economic Review 56 (1966). For more recent literature parting from the view that PuEEMME enterprises seek to maximize profits per employee see: John Bonin, "The Theory of the Labor Managed Finn from the Membership's Perspective with Implications for Marshallian Industry Supply," Journal of Comparative Economics 5, no. 4 (December 1981): 337-5 1; John Bonin and Luis Putterman, "Economies

of Cooperation and the Labor Managed Economy," in Fundamentals of Pure and Applied Economics and the Encyclopedia of Economics, J. Lesourne and H. Sonnenschein, eds. (1985); Norman Ireland and Peter Law, The Economics of Labor-Managed Enterprises (New York: St. Martins Press, 1982); and J. M. Montias, "On the Labor-Managed Firm in a Competitive Environment," Journal of Comparative Economics 10, no. 1 (1986): 2-8.

8. Again, we note that critics also charge that PuEEMMEs are less able than PrEMEs to adjust to new general equilibria when technologies and preferences change. But since we have focused on the properties of market equilibria and deliberately left issues of dynamic adjustment aside, we will not concern ourselves with this side of the debate.

9. This is true for the same reason enterprises in Debreuvian models of competitive PrEMEs cannot be expected to achieve equal rates of profit on "capital advanced." What Debreu calls "profits" are actually rents or "quasi rents" that are the result of the unique technological capabilities of a particular enterprise, which are not competed away because each firm is distinguished by its unique production possibility set and the number of firms is fixed in his model.

10. Gerard Debreu, Theory of Value (New York: Wiley, 1959), chap. 6. More precisely in his model and analysis

1. For PuEEMME the definition of an equilibrium (Debreu, sec. 6.2) must be changed to read: "yj* maximizes pyj/yh on Yj, for every j where h is the input labor" instead of "yj* maximizes pyj on Yj for every j" (p. 93).

2. In the proof that an equilibrium is an optimum (Debreu, sec. 6.3), although it is still true in PuEEMME that " xi* minimizes pa" on Xi," it is no longer the case that "-yj* minimizes Pa on -Yj" so the conclusion

optimum (Debreu, 94)-no longer follows. 3. In the proof that an optimum is an equilibrium (Debreu, see. 6.4), an optimum allocation still requires that ^yj is such that"-^yj minimizes Pa on -Yj for every j." But it is not true that -^yj is the same as -yj* the equilibrium production such that yj* maximizes pyj # yh on Yj.

11. Branko Horvat, Towards a Theory of Planned Economy (Belgrade: Yugoslav Institute of Economic Research, 1964); idem, "Critical notes on the Theory of the Labor Managed Firm and Some Macroeconomic Implications," Komunikacije (1972); and Jan Vanek, The Economics of Workers' Management: A Yugoslav Case Study (London: Allen and Unwin, 1972).

12. Horvat, "Critical notes," 290.

13. Vanek, Economics of Workers' Management, 174-76.

14. Branko Horvat, "The Theory of the Worker Managed Firm Revisited," Journal of Comparative Economics 10, no. I (1986): 12.

15. Jaroslav Vanek, The General Theory of Labor Self-Managed Economies (Ithaca, N.Y.: Cornell University Press, 1970),135.

16. Dilip Madan, "Long-run Competitive Equilibrium and the Capital Controversy," Economics Discussion Paper no. 3174, (Bundoora Victoria, Australia: Department of Economics, La Trobe University, April 1974). William Novshek and Hugo Sonnenschein, "General Equilibrium with Free Entry: A Synthetic Approach to the Theory of Perfect Competition," Journal of Economic Literature 25, no. 3 (September 1987): 1281-1306, published a similar model.

17. Of course, in perfectly competitive PrEMEs there are no monopsonistic input markets. But if any such markets should become less than perfectly competitive, it is well known that a private enterprise market economy fails to produce socially efficient outcomes. PuESMMEs of the Lange-Lerner-Taylor variety are not subject to this problem even should input market structures be monopsonistic.

18. Again, perfectly competitive PrEMEs do not have monopolistic product markets. But should a market become monopolistic in private enterprise market economies inefficiency results, whereas in PuESMMEs it would not.

19. In any case, the "pseudomarket" adjustment mechanism, which was inexplicably proposed only for intermediate goods, is not a necessary feature of PuESMMEs. One could postulate a PuESMME with "freemarket" adjustments everywhere, or with pseudomarket adjustments everywhere, for that matter. In any case, this issue affects dynamic adjustment and stability of equilibria, but does not affect the optimality properties of equilibria once they are achieved.

20. It should be noted that losses are not necessarily a sign of lack of effort or inefficiency in Lange-Lerner-Taylor enterprises. It is possible that losses are precisely the result of obeying the managerial rules that are substituted for the criterion of profit maximization.

21. Under Lange's bonus distribution rule, all workers would be induced to work too much and take too little leisure time given their actual preferences for income versus leisure.

22. PuESMMEs do not fare nearly so well in the eyes of our new approach. PuESMMEs share with PrEMEs the dubious distinction of being doubly inefficient according to our analysis, whereas PuEEMMEs and PuECPEs display snowballing inefficiency along only a single axis. But this argument is yet to come.

23. As the reader likely suspects, there are reasons for expecting such biases in both PuEEMME and PuESMME according to our new welfare paradigm, as we will see shortly.

24. Marx's famous "general law of capitalist accumulation" is the best known of such theories. But there has been a resurgence of interest among modern political economists in "the political economy of the business cycle." For example, see: Rayford Boddy and James Crotty, "Class Conflict and MacroPolicy," Review of Radical Political Economics 7 no. I (Spring 1975): 1-17; Thomas Weisskopf, "Marxist Crisis Theory and the Rate of Profit in the Postwar U.S. Economy," Cambridge Journal of Economics 3 no. 4 (December 1979): 341-78; Robin Hahnel and Howard Sherman, "Income Distribution Over the Business Cycle," Journal of Economic Issues 13 (March 1982):49-73; Samuel Bowles and Herb Gintis, "The Crisis of Liberal Democratic Capitalism," Politics and Society 11 no. 1 (1982): 69-79; and Howard Sherman, "Changes in the Character of the U.S. Business Cycle," Review of Radical Political Economics 18, nos. 1 and 2 (Summer 1986): 190-204.

25. The development of post-Keynesian disequilibrium theory is one example. See Axel Leijonhufvud, On Keynesian Economics and the Economics of Keynes (New York: Oxford University Press, 1968); and Robert Clower, Monetary Theory: Selected Readings (Harmondsworth: Penguin, 1969). For a brief, but excellent summary presentation see Laurence Harris, Monetary Theory (New York: McGraw Hill, 1981), chap. 13.

26. One of the foremost scholars of PuEEMMEs published a whole book on the subject. Branko Horvat, Business Cycles in Yugoslavia (White Plains: International Arts and Sciences Press, 197 1).

27. Such is the case for Lange, Taylor, and Lerner's version of PuESMME, as we pointed out above.

28. See Paul Blumberg, Industrial Democracy: The Sociology of Participation (New York: Schocken Books, 1969).

29. See Maurice Dobb, Soviet Economic Development Since 1917 (New York: International Publishers, 1968).

30. In a similar vein, the ability of producers in a single unit to make sensible choices about their activity depends on their knowledge of their productive possibilities. But if one does not envision workers deciding how to organize their own activities, but instead obeying instructions from an owner or manager, then, obviously, it is the owner or manager who must be aware of the possibilities in order to make informed decisions.

31. We have obviously arrived at the traditional view of markets as cybernetic miracles that dates back at least as far as Adam Smith. Markets

eliminate the need for anyone to receive information about what is, without doubt, very extensive and complicated interrelations. As long as actors understand their individual goals and possibilities they need not know anything about others' situations for a market system to function effectively. The miracle is not so much the information that markets supply since they supply precious little information at all. The miracle is rather that actors need not know anything about others to make their choices in market economies. As a matter of fact, providing everyone with as much information about others as they have about themselves would presumedly not change the decisions they make one iota in a competitive market economy.

32. Although there are important differences between barter and monetized exchange systems, they are irrelevant to our purposes and we can safely ignore the issue.

33. Paul Sweezy provides a typical Marxist formulation of this informationdisguising property of markets: "The quantitative relation between things... is in reality only an outward form of the social relation between the commodity owners.... The exchange relation as such apart from any consideration of the quantities involved, is an expression of the fact that individual producers, each working in isolation, are in fact working for each other. Their labor, whatever they may think of the matter, has a social character, which is impressed upon it by the fact of exchange." Quoted in R. Edwards, M. Reich, and T. Weisskopf, eds., The Capitalist System, 1st ed. (Englewood Cliffs, N.J.: Prentice Hall, 1972), 112.

34. Indeed, as we will see shortly, Arthur DiQuattro suggests something along these lines as a way of avoiding "commodity fetishism" in PuEMEs.

35. Deleting such information when decisions are to be made by profitmaximizing owners or managers who can have no use for such details is obviously economizing.

36. Arthur DiQuattro, "Alienation and Justice in the Market," American Political Science Review 72, no. 3 (September, 1978): 885-86.

37. To his credit DiQuattro notes that Karl Marx did not subscribe to his view. DiQuattro quotes Marx: "Goods produced for exchange by workers who do not come into social contact with each other until they exchange their products are bound to assume a fetishistic quality in the experience of the producers" and "Fetishism... attaches itself to the products of labor as soon as they are produced as commodities, and which is therefore inseparable from the production of commodities" (DiQuattro, "Alienation and Justice in the Market," 885).

38. Karl Marx, Economic and Philosophical Manuscripts of 1844 (New York: International Publishers, 1964), 148.

39. Adam Smith, The Wealth of Nations (New York: Random House, 1937),14.

40. Letter from Karl Marx to P. V. Annekov, December 28, 1846, in Karl Marx; Collected Writings (New York: International Publishers, 1940).

41. Gar Alperowitz, "Socialism and a Pluralist Commonwealth," in The Capitalist System, Edwards et al., eds., 5 27.

42. Frank Roosevelt, "Market Socialism: A Humane Economy?" Journal of Economic Issues 3, no. 4 (December 1969): 18.

43. Different forms of taxation and expenditure policies as well as voluntary associations are included here under the title "makeshift social structures."

44. E. K. Hunt, "A Radical Critique of Welfare Economics," in Growth, Profits, and Property, E. Nell, ed. (Cambridge: Cambridge University Press, 1980), 244.

45. The increasingly prevalent belief among defenders of market systems that external effects can be adequately handled by an appropriate (re)definition of individual property rights completely misses the point. All that is accomplished by a redefinition of individual property rights is a swapping of who has the right to generate external effects for whom. No redefinition of individual property rights can ever move us from a situation with external effects to a system without them.

46. We defined utility functions on individual consumption and work space in chapter 6, only because the points we wished to demonstrate in that chapterthe implications of endogenous preferences--could be proved under the traditional assumption of external effect exceptionality. However, as the paradigm developed in chapter 5 made clear, we believe individual well-being functions should be defined on all human activity space,

47. K. J. Arrow, "Political and Economic Evaluation of Social Effects and Externalities," in Frontiers of Quantitative Economics, Michael D. Intriligator and D. A. Kendrick, eds. (Amsterdam: North Holland Publishing, 1974),18.

48. Notice that "guide" is not the same as "solution." While the concept of self-management may help clarify the nature of what we strive for, it does not tell us how best to achieve such a decision-making arrangement. There is not even any guarantee that such an arrangement is possible.

49. There is no clear-cut goal in terms of the traditional paradigm for two reasons. When one person's individual freedom conflicts with another's in the sense that its exercise would violate the individual freedom of another, there is no guide to action. While some might claim that this is precisely the role, played by property rights in the traditional paradigm, property rights only establish a clear guide if there are justifications for one system of property rights over another. But the second ambiguity arises from the

conflict between the imperative of free exercise of individual freedom with the imperative of social efficiency, which is how Arrow sees the dilemma in the quote above.

50. Again, the important "dilemma" of finding a system of economic decision making that accomplishes this aim is not a dilemma of criteriathat is of our welfare theory-but a dilemma of economic system design.

5 1. Hunt, "Radical Critique," 24 1.

52. Ibid., 245.

53. Richard Musgrave, The Theory of Public Finance (New York: McGraw Hill, 1959),134.

54. Arrow, "Political and Economic Evaluation."

55. Ibid., 16.

56. Ibid., 17.

57. As a matter of fact, there is no way for the coalition of affected partners to effectively challenge claims regarding coalition membership. If the coalition is negotiating a bribe to be paid to the actor creating the external effect, it is in everyone's interest to decline membership even if they are affected. If the coalition is negotiating a bribe to be received from an actor who must obtain coalition approval to do what he or she wishes, it is in everyone's interest to claim membership even if they are unaffected.

58. Note that the only role "property rights" plays is in determining whether or not it is the coalition of affected parties that must bribe the economic actors whose activity affects them, or the economic actors who must pay the affected coalition for infringing on their "right" to be unaffected. Property rights merely establish whose "individual freedom" takes precedent over whose. If actors have the legal right to do as they please, it is the coalition of affected parties who must bribe in order to influence the decision. If the affected parties have the legal right not to be affected, actors must bribe the coalition in order to obtain their permission to do what the actors please. In either case, the dilemma we are explaining is the same. Coalition members have an incentive to misrepresent the degree to which they are affected to minimize their share of the assessment or maximize their share of the payment received, and there is no effective means for the coalition to challenge such misrepresentations without cutting off their noses to spite their faces.

59. In Arrow's opinion the argument that unrestricted bargaining will lead to Pareto efficient outcomes "is not easy to evaluate in the absence of a generally accepted concept of solution for game theory," despite arguments to the contrary by those such as Zeuthen, Harsanyi, Shapley, and Selten. And since Von Neumann and Morgenstern generalized Edgeworth's nineteenth-century conclusion that the results of such bargaining are in determinate, it is not difficult to foresee nonoptimal outcomes. In Arrow's words, "it is certainly a matter of common observation, perhaps especially in the field of international relations, that mutually advantageous agreements are not arrived at because each party is seeking to engross as much as possible of the common gain for itself." In a similar vein Arrow observes that "each owner of a small parcel whose acquisition is essential to the execution of the enterprise can demand the entire net benefit [in which case] an agreement may never be reached or may be long delayed. At positive discount rates even the latter outcome is not Pareto efficient. It is to avoid such losses that the coercive powers of the state are invoked by condemnation proceedings" (Arrow, "Political and Economic Evaluation," 10).

60. To keep matters tidy we would have to assume that no individual's use of the park or bridge diminishes its usefulness to others and that ships pass by the lighthouse one at a time.

61. Arrow, "Political and Economic Evaluation," 17.

62. Note that this conception already assumes interpersonal utility comparisons. Each individual reports marginal willingness to pay in terms of some moneylike good.

63. Musgrave, Theory of Public Finance, 134.

64. It would not matter whether taxes were proportional share, based on income, wealth, or any other criterion. As long as citizens who truly care little for the public goods are "compelled" to pay a positive tax, they would be worse off than under a market/voluntary association allocation.

65. Note that a citizen who benefits from the public goods would be made worse off by a move from a system with government back to free-market anarchy, implying that this move cannot be a Pareto improvement either. But this does not mean there is no difference between the government that collects taxes unrelated to benefits and free-market anarchy with regard to efficiency. If the government provides public goods to equate marginal social benefits and costs, it will achieve Pareto optimality irrespective of how it decides to tax people, The taxes are simply lump-sum redistributive transfers affecting equity. Free-market anarchy, on the other hand, cannot achieve Pareto optimality if there are external effects.

66. Musgrave, Theory of Public Finance, 133-34.

67. P. A. Samuelson, "The Pure Theory of Public Expenditures," Review of Economic Statistics 36 (November 1954): 387-89.

68. C. M. Tiebout, "A Pure Theory of Local Expenditures," Journal of Political Economics 64, no. 5 (October 1956): 416-23.

69. We interpret property, wealth, or income taxes here as an indication of "ability to pay" rather than as an indication of the degree to which one is

benefited by the maintenance of public order responsible for preserving the integrity of one's property, wealth, or income.

70. This need not be the case if the government chooses to ignore citizens' preferences in deciding how much and what kind of public goods to provide. But we would think in any system with claims to efficiency and equity citizens' preferences would be taken into consideration and individuals' preferences given more or less equal weight.

7 1. We emphasize that we are concerned here with issues of central finance, so that providing minorities with different packages, implementing more than one solution, and "voting with one's feet" are ruled out. In central finance, satisfying minority preferences implies denial of majority sentiments.

72. Again, we assume individuals' preferences have more or less equal weight in determining what public goods will be provided.

73. See Samuelson, "Pure Theory of Public Expenditures," 387.

74. We should note that the snowballing nonoptimality will be all the worse to the extent there is a correlation between the sign and degree of public characteristics themselves and the developmental effects of consuming such goods on future preferences. In other words, while there will be snowballing nonoptimality in any event, if engaging in social and/or private activities is "habit forming" in Pollak's sense, the snowballing nonoptimality will be all the worse.


1. It is of course skepticism regarding the likelihood of such repeated successful applications of political will that leads egalitarian critics of PrEMEs to challenge its practical claim to fairness.

2. In a Debreuvian treatment of time this is not an issue because all choices for all periods are made at once with perfect foresight-before anything starts happening. That is why "lump-surn redistribution" does not pose incentive problems in a Debreuvian world. It only happens once, and before people begin to act at that. The third fundamental welfare theorem says any Pareto optimum can result from a private enterprise market economy under the appropriate redistribution. It should be understood that Debreuvian "redistributions" reshuffle endowments for all actors in all time periodsonce, before the economy begins to operate. And in Debreuvian models there are no unforeseen inequalities that accumulate and need to be ironed out over and over again along the way. All eventualities were accounted for in that single, all-seeing redistribution before things got started. But while this could happen in a Debreuvian world without influencing incen

tives, the third fundamental welfare theorem does not tell us a particular Pareto optimum can be achieved by repeatedly redistributing along a trajectory through real time. Conservatives are correct to point out this is a misconception of the flexibility theorem that ignores the obvious incentive implications of such a policy.

3. There are always exceptions. The stability of egalitarian redistribution in Sweden, for example, is impressive. But we suspect that comparison of the degree of inequality in private and public enterprise economies on average would bear out the view that expecting egalitarian distributions from private enterprise systems is like swimming against the current: possible, but not the easiest way to cover distance through water.

4. This is not the case for human capital theory, but then human capital theory had to overcome the myopia of the traditional paradigm to which we refer before being accepted as part of traditional theory.

5. Even here there are subtleties to be considered. To the extent that human productive assets are generated through training or education, while not exactly "transferable," at least individuals that begin without them can acquire them if we are willing to wait the necessary time. Cognizance of the time that may be necessary certainly affects the "meaning" one attaches to the third fundamental theorem. But more importantly, there are presumably some human productive assets which are not the product of training or education, but due to genetic endowment. To the extent that human assets that the combination of preferences, income distribution, technology, and institutional influences render "productive" to differing extent are of this nature, there is no meaningful sense in which they could be considered transferable short of a tampered rerun of the genetic lottery!

6. Again, we are not speaking of repeated assessments that would necessarily affect incentives, but of a one-time-only assessment for all periods of time that occurs with perfect foresight prior to the first period.

7. Indeed, the very same "practical" considerations that lead opponents of PrEMEs to be unswayed by the theoretical income flexibility of PrEMEs, and insist that PrEMEs will display what they view as income distribution "problems," operate in PuEMEs to militate against redistributions those who favor PuEMEs would find objectionable.

8. Alternatively we could pay people their salary whether they are working, in training, or in formal educational programs.

9. If asked we expect most would testify that their "school" time had been more enjoyable than their "work" time. But those who receive a greater than average share of education are usually as quick to forget this fact as they were happy to benefit from it. In any case, in public enterprise economies the fiscal assumptions are largely fulfilled, and education is "public" to a great extent in many private enterprise economies as well.

10. No doubt Nozick and others attracted to this maxim would object to our formulation. They would probably insist on something like: from each according to his or her genetic and property inheritance, to each according to his or her genetic and property inheritance; or better still, from each according to his or her productive assets, to each according to the full contribution of his or her productive assets. And by their assets Nozick and followers would mean both their personal abilities and the potential contribution of the nonhuman assets they own. We do not wish to quibble, but think there is a meaningful sense in which the use of someone's human capabilities must come "from" that person, whereas the use of physical objects requires that nothing come "from" those deemed their owners.

11. Of course, the more usual phrasing is: (3') from each according to ability, to each according to work, and (4') from each according to ability, to each according to need. But the interpretation of 3' has been the subject of great controversy among Marxists and all concerned with "postcapitalist" economies. If one means by "work" what we call loosely "effort," meaning some appropriate combination of time and intensity of labor along with degree of disagreeableness--then 3' is merely another wording of 3. But if one means by "work" "the productivity of one's efforts as determined by the conditions of the economy and one's genetic and accumulated human assets," as those we have described elsewhere as "coordinator apologists" interpret 3" then 3' is another wording of maxim 2, and should be relabeled 2'. While there has never been much doubt about the meaning of 4" traditionally Marxism has viewed increases in economic productivity to the point of nonscarcity as the key to its implementation. We regard this self-delusion at best, and at worst an attempt to deflect rightful impatience and resentment on the part of those disadvantaged during the supposed "transition." For both ecological and sociological reasons the traditional Marxist vision of how economies might arrive at 4 is misleading. In our view, the key to a "transition" to 4 is a lengthy period of solidarity building in 3.

12. Which is not to say powerful political pressure may not be mounted by those with greater "human capital" in PuEMEs to reprivatize payments to nonhuman productive assets, as can be observed in China and the Soviet Union today. After all, the fact that PuEMEs place limits on individuals' accumulation of wealth and their ability to pass on wealth to their heirs will not go unchallenged by those who deem themselves "fettered."

13. In John Roemer, A General Theory of Exploitation and Class (Cambridge, Mass.: Harvard University Press, 1982), he effectively defines "capitalist" exploitation as income distributions that express Maxim 1, and "socialist" exploitation as income distributions that express Maxim 2. In his terms we find "socialist" exploitation no more justifiable than "capitalist" exploitation. In fact, in societies created by movements

seeking social justice, the historical rather than purely ethical imperative to eliminate inequalities is even stronger in our view! But because socialist movements have historically sought economic justice as one of their principal goals, we reject Roemer's terminology as inappropriate. For us Maxim 3 corresponds to economic justice, fairness, or the absence of economic exploitation---one of the principal historical goals of socialist movements. And income distributions expressive of Maxim 3 would not contain what Roemer calls "socialist" exploitation. The undeniable fact that such exploitation exists in societies that are commonly called "socialist" is simply evidence that they fail to correspond to the societies envisioned by the socialist movements that may have inspired them, and that such societies are labeled inappropriately. We would remind those tempted to accuse us of confusing socialism with communism that Maxim 3 is quite different from Maxim 4, which corresponds to human solidarity-a social relation beyond fairness. Which is not to say that the trust created by a lengthy history of fairness is not the most likely "transition" to solidarity. We might add that recently John Roemer has offered a persuasive argument against the morality of what he terms "socialist exploitation" as well as "capitalist exploitation." See John Roemer, Free to Lose (Cambridge, Mass.: Harvard University Press, 1988), chaps. 5, 9, and 10; idem, "Equality of Talent," Economics and Philosophy 1 (Fall 1986); and idem, "Equality of Resources Implies Equality of Welfare," Quarterly Journal of Economics 101 no. 4 (November 1986): 75184.

14. Ariel Rubinstein, "Perfect Equilibrium in a Bargaining Model," Econometrica 50 (1982): 97-110; and Gil Skillman, Jr., "Sequential Bargaining in Capitalist Firms," working paper no. 88-28, Department of Economics, Brown University (November 1988).

15. For example, in the 1970s it was well known in the Washington, D.C., area that private consulting firms working on defense projects had to pay a "wage premium" for professional workers in comparison to firms that consulted only on civilian contracts.

16. Until the recent work of John Roemer in A General Theory of Exploitation and Class, who would be employers and who would be employees was determined exogenously in Marxist analyses. In most cases it was simply assumed that some small group of individuals owned all the means of production and everyone else was free only in the sense that they were free to work for whichever of these employers would have them. Roemer's endogenous determination of "class" position based only on the assumption of an unequal initial distribution of wealth is unique in the Marxist approach to this matter. Roemer's work in this regard is provocative and illuminating, the frenzied rejection from Marxist officialdom notwithstanding.

17. After all, theories of capitalist "crisis," such as the theory of the "tendency for the rate of profit to fall" and "under consumption" theories, which have long enamored Marxists, would merit no attention if individual capitalists could always be counted on to behave in their class' best interests. We say this without implying that after careful consideration most of these theories merit much attention in any case. Our point is simply that if capitalists were assumed to be capable of conspiring to always act in their class' interests, such crises would be prima facie impossible.

18. We do not reproduce Reich's formal demonstration of this result since we will re-derive this result and others in a more general model below. See Michael Reich, "Models of Class Conflict and Racial Inequality Within the Firm" in Racial Inequality (Princeton, N.J.: Princeton University Press, 1981), 204-15, appendix to chap. 5, for Reich's demonstration.

19. We are suggesting here that traditional theorists might "certify" Reich's criticism as theoretically unassailable, but dismiss it as insignificant in practical terms. This would be similar to the rebuttal Keynesians offered to Pigou's argument that falling prices would raise real assets and, thereby, raise private consumption, eventually lifting economies out of "liquidity traps." Keynesians "certified" Pigou's logic, but challenged the empirical significance of the "Pigou Effect," arguing that there would be many red faces if we all held our breath and waited for the "Pigou Effect" to save the day.

20. When we use the term "real wage," we mean payment per unit of work effort, so the "real wage" incorporates both payment and "extraction." For example, the real wage would fall if payment declined or extraction increased. Since inflation is not a phenomenon we treat, there should be no confusion with the other (more common) sense in which "real wage" is used.

21. None of which would be any surprise to previous contributors to the conflict theory, nor to most traditional theorists.

22. We do not mean to rule out the possibility implicit in Reich's analysis that present work activity can affect present characteristics. But in our view, it is far more plausible that important effects would be felt in the future, In any case there are many more periods in the future than in the present, so the total future effects are likely to be greater than the present effects.

23. In another sense the human characteristic transforming effects of work activity matter to employees. Since characteristics "parameterize" preferences, work activity has a preference-developing effect as well as a preference-fulfilling affect. While traditional theory is blind to the preferencedeveloping effects of economic activities in general, we can easily lump these effects together with the preference-fulfilling effects commonly discussed as job satisfaction--or lack thereof. We do not

emphasize this aspect of the situation here because, as with preference fulfillment effects, employers have no reason to care one way or another about these effects. Except for the fact that workers may be willing to accept lower wages for greater preference-developing effects, employers have no interest in the matter. The assumption of competitive labor markets can be interpreted as guaranteeing compensating market wage differentials for these effects, but in any case employers have no "axes" of their own "to grind."

24. The reader should not confuse a degree of endogeneity with a degree of indeterminacy. As we will see, perfect information, competitive models permit no indeterminacy in wage or effort determination. But part of the "determinate" outcome is an endogenously determined wage-effort effect that derives from employer manipulation of conditions he or she controls in the production process. Moreover, we demonstrate that this "determinate" outcome is socially inefficient.

25. So far these are rebuttals to other's formulation of the "conflict theory" of the labor process under private ownership.

26. In many, if not all, actual, historical, private enterprise settings there are a number of deficiencies that our generous assumptions have permitted to pass unchallenged. We had every intention of organizing the "debate" in this way, but take this opportunity to remind the reader of the rarified theoretical world we are traveling in!

27. Traditional theory confirms this interpretation in the form of the zero profit assumption in competitive, steady state models.

28. Again, this is consistent with traditional assertions to the same effect. Interestingly enough, Marxist authors such as John Roemer who abstract from the difference between labor power and labor and stipulate competitive labor markets come to the same conclusion. The isomorphism between Roemer's "Labor Market Island," in which those with financial capital hire wage labor, and "Credit Market Island," in which those with financial capital lend it to those without, is an excellent example of the effect of abstracting from one of the two distinguishing features of what most call capitalismnamely the institution of private ownership and management of the means of production. See Roemer's General Theory of Exploitation and Class.

29. While it is not relevant to the present argument, we seriously doubt that if employees ever even approximated this degree of power over employers they would "choose" to maintain private enterprise market institutions. At a minimum there would be every incentive to opt for an employeemanaged, public enterprise market economy of the kind we discussed in the preceding chapter. And, in our opinion, once they understood the deficiencies of market institutions they would opt for some system of participatory planning in conjunction with public enterprise.

30. Whether we have "homogeneous" labor of a single category and productivity or "heterogeneous" labor of different categories and/or productivities is irrelevant. When we talk of "net product" that under one set of conditions would go entirely to employees and under another set of conditions would go entirely to employers, this "net product" could be divided up among one, or more than one, category of employees, who could receive the same or quite different wage rates.

31. Admittedly, this is not a particularly "traditional" view of financial capital. Ultimately financial capital is simply credit-the willingness of some agents in the economy to trust other agents' promises of repayment. In this light the "supply" of financial capital is not a fixed number but a supply schedule with a degree of elasticity. Put this way, our proposition becomes that competitive credit markets (which are merely the "flip side" of competitive labor markets) do not guarantee an infinitely elastic supply of credit at a rate of interest that is only infinitesimally positive. Traditional theorists no doubt would object there is no reason to believe credit would not be infinitely elastic at an infinitesimal interest rate in a Steady-state world inhabited by immortals rendered risk free by the full knowledge assumption. Perhaps, but can't we then be excused for considering a world in which some put others to work without expectation of any return for themselves somewhat "rarified"? Notice that whether the credit derives from employers' own wherewithal or from others who finance "entrepreneurs" seeking to expand activities with positive returns beyond the limits of their own wherewithal is of no consequence. It is still a matter of whether some agents in the economy are willing to put others to work with no expectation of return for themselves.

32. Admittedly, it is possible to rephrase this conclusion in a way that appears to preserve the conclusions of traditional theory. If the zero rate of profit assumption is taken as a proxy for complete domination of employee over employer with regard to the conflicts of interest between them, it is possible to continue to insist that under traditional assumptions, which include a zero long-run rate of profit, private employers' choice of technology and reward structures will be Pareto optimal and nondiscriminatory. But the conclusion can now be seen for the trivial tautology it is. Once a zero rate of profit is seen to stand for complete labor dominance, it is trivial that employers cannot do anything other than what their employees would wish them to do! Of course we have "producer sovereignty" -all powerful employees would be irrational to permit anything less. Similarly, it would be irrational of all-powerful and all-knowing employees to permit themselves to be manipulated by discriminatory ploys into a situation in which they have less than complete control.

33. To see this, imagine an economy in which there are millions of employees and millions of employers, complete mobility of both labor and

capital, and job information available to all at electronic speeds-i.e., indisputably competitive labor markets. If employees everywhere were characterized by complete knowledge of production procedures, full solidarity with one another, a strong conviction that only those who actually work deserve reward, and a great amount of energy and militancy for pursuing their convictions, the normal rate of profit might well approach zero. On the other hand, if employees are largely ignorant of how to carry out production without direction and supervision, suspicious and antagonistic toward one another, receptive to employers' claims to substantial rewards, and totally devoid of militancy, it is inconceivable that the normal rate of profit in the economy would not be significantly positive.

34. The choice of material input and output mixes, a principal concern of traditional theory, is one obvious example where nothing is lost by abstracting from the strength of employers' bargaining positions vis-a-vis their employees.

35. Presumably, few would argue this assumption is without interest given the history of "real world" private enterprise economies!

36. One might conclude from this that traditional theorists interested in preserving the accuracy of their conclusions would be honor-bound to root for employees to consolidate their power over employers, for otherwise "producer sovereignty" and "productivity-based wages" will not obtain. At a minimum, they should point out the illogic of highly prominent claims to the contrary, namely that reducing labor's bargaining strength will improve the "efficiency" of the U.S. economy!

37. It does not concern us if it is the "immediate" employer who earns positive profits or an "indirect employer" in the form of a financial backer who receives the positive profits as interest payments from the "immediate" employer.

38. Which is not to say we believe employees would be best off doing nothing, if they could get away with it, or that employees could not gain positive satisfaction from work efforts they knew to be socially productive and fairly distributed.

39. We are not speaking of Gintis and Katzner here, since their point was precisely the implausibility of their sufficient conditions.

40. Not all public enterprise, employee-managed economies are the same. For present purposes the important features are: (1) The employees are empowered to manage the enterprise either directly or through delegates of their choosing, and (2) the well-being of every employee is directly affected by the performance of his or her fellow employees. The latter might be because the performance of fellow employees affects one's income or because it affects the burdensomeness of one's work assignments. Both public enterprise, employee-managed, market economies (PuEEMMEs), and public enterprise, employee-managed, decentrally planned economics (PuEMDPEs) meet these conditions. While Bowles favors the former and we favor the latter, the differences are not important to the issue at hand.

41. Fellow Productive employees might also be in a better position to detect malfeasance than supervisory employees who are not engaged in productive activity. But this effect implies asymmetric knowledge about productive possibilities on the part of employers (or their supervisors) and employees. And while we believe this assumption is frequently reasonable, we also believe, as did Bowles, that there is an argument for inefficiency independent of asymmetric knowledge of production. Therefore, we waive the point for the same methodological reasons Bowles did.

42. By "strictly welfare theoretic terms," we mean explanations resting entirely on the assumption of pursuit of rational, individual self-interest under particular circumstances, without implying that actual human behavior is not considerably more complex.

43. Translated into traditional terms, empathetic preferences are characterized by positive direct externalities am happier if you are happier.

44. Presumably most would agree it is not irrational, for instance, to neglect the interests of one's employer more than the interests of one's family and friends even if there is greater opportunity to shirk in the latter case. Our paradigm and concepts help explain why this is the case. For a formal demonstration of the argument in terms of the new paradigm see the last two sections of this chapter.

45. For a fuller discussion of these issues see Michael Albert et al., Liberating Theory (Boston: South End Press, 1986).

46. To his credit Reich notes that both possibilities exist within his model, But Reich takes a strong stand that in the United States, white workers have lost more than they have gained from racial discrimination. For someone who argues the opposite case, see Michael J. Carter, "Competition and Segmentation in Internal Labor Markets, "Journal of Economic Issues 15 (December 1982): 1063-1077. Unlike Reich, Carter errs in thinking the case for net gains to white employees can be proved theoretically. On the other hand, we do not find Reich's empirical evidence to the contrary conclusive and retain an open mind on the balance of effects in the U.S.

case. But the actual outcome during a particular period in the U.S. is neither here nor there as far as the general theory of such relations is concerned.

47. For an excellent assessment of the evidence with respect to economic discrimination against women in the U.S., see the series published by National Academy of the Sciences Press, Washington, D.C.: Donald Trieman and Heidi Hartman, eds., Women, Work, and Wages (1981); Barbara Reskin, ed., Sex Segregation in the Workplace: Trends, Explanations, Remedies (1984); Heidi Hartman, ed., Comparable Worth: New Directions for Research (1985); and Barbara Reskin and Heidi Hartman, eds., Women's Work, Men's Work: Sex Segregation on the Job (1986).

48. There are additional reasons, in the real world, why racial and sexual discrimination rooted in spheres of social life outside the economy will necessarily bias economic relations as well. See Albert et al., Liberating Theory for discussion and references.

49. If those released remain unemployed, and the quantity or quality of outputs resulting from less skill-intensive or less skill-dispersed technologies declines, there is, of course, a loss from society's point of view.

50. For private employer "economizing" of scarce productive skills to be fully consistent with the social interest, the relative wages of different categories of labor must obviously be accurate measures of relative scarcities and job satisfactions. While in the argument that follows we waive the point, there is no reason to believe the kinds of biases we have described in market prices do not apply to wages as well. At a minimum, wages must be indirectly affected by biases in other prices. And the discussion to come will make clear that in private enterprise economics there is every reason to expect factors other than productivities and desirabilities enter into wage determination in any case.

5 1. In this light, the incentive for employers to engage in non-job-specific, onthe-job training resides in the absence of labor markets sufficiently competitive and "tight" to generate full compensating wage differentials for more skilled employees who remain with the employer who "trained" them. Alternatively, such programs are means of attracting better employees, but in this status they are simply one among many "come-ons."

52. It is possible for excessive fragmentation to diminish productivity, as many studies of quality control problems indicate. This does not negate our point in the least. It simply means that in some situations a profit-maximizing employer may have to weigh the trade-off between the profit-increasing "bargaining power" effect and the profit-decreasing "productivity" effect of increased fragmentation.

53. One would have to examine what circumstances would generate such a condition. But seniority rules, firm-specific skills, and pension contribution/benefit structures are among a number of possibilities that have been explored.

54. This latter effect was ignored in Reich's model, but as we explained, employer bargaining power is just as likely to affect wages as effort extracted. We remind those who might think we have contradicted our assumption of a competitively determined wage rate by making wages partially dependent on employer bargaining power of our previous argument that the assumption of "competitive labor markets" does not settle the issue of wage and effort determination. The entire discussion is premised on the assumption of some degree of endogeneity in the deter 0 initiation of those issues over which employers and employees have conflicting interests. If there is any endogeneity-if the competitive labor market only establishes limits on the wage/effort package that results-it is just as appropriate to recognize the potential effects of employer bargaining power on wages paid as on effort contracted. After all, if there is no degree of endogeneity, there is no possibility of employer bargaining power affecting effort since in this case the competitive labor market would presumably determine not only the wage rate, Wk, but the degree of effort that would expended per hour as well. It is simply that we are not used to writing down that part in traditional notation. Put differently, any who would object to stipulating the wage rate as a partial function of employer bargaining power should object to writing effort as a function of bargaining power as well.

55. It is clear that the nontraditional future wage effect (wage enhancing for work type-1, wage reducing for work type-2) has only a distributional impact. Nothing more is produced and no additional effort or time is expended. But the nontraditional effort effect does affect future output. However, this is not a "social benefit" because the increased output from additional effort is the result of increased effort from employees. While increased effort is of no consequence to employers who do not have to pay for it, it must be added to social costs. If we can assume that competitive labor markets have succeeded in equating the marginal disutility of work done with the wage paid, the increased future output is exactly canceled by the increased future disutility of the extra effort to be expended.

56. Since the rate of labor turnover varies dramatically in different private enterprise economies around the world and between different industries within those economies, there are obvious possibilities for empirically testing the strength of this effect.

57. "Among employees" is important here, since the argument is that perceptions of fair treatment of employees vis-a-vis one another, or the absence of unfair "privileges" among employees, contributes to employee solidarity. This is not inconsistent with the claim that employee perceptions ,of unfair distribution of duties and rewards between owners and employees contributes to employee solidarity as well.

58. The discrimination may be on the basis of race, gender, sexual preference, or any other social characteristic that generates suspicion, mistrust, and hostility.

59. The phrasing in parentheses is the more common terminology of those who think in terms of the conflict school. We supply the translation for those unfamiliar with either semantic.

60. The sum of terms in the second line of expressions (8. 1) and (8.2), the nontraditional effects, are precisely what summarizes the effects of such practices on employer profits.

61. The proof is immediate applying Theorem 6.6 to Theorem 8.1 written for period (t+k).

62. Obviously these are not exactly the "most favorable institutional conditions imaginable," since we do not conclude that PrEMEs are cybernetic miracles and efficiency machines.


1. Enrico Barone, "Ministry of Production," in Collectivist Economic Planning, F. A. Hayek, ed. (London: Routledge Press, 1935).

2 Later, we will analyze procedures by which planners might accumulate such information.

3. A number of texts elaborate the theory and various solution algorithms of different kinds of mathematical programming problems. See G. Hadley, Linear Programming (Reading, Mass.: Addison Wesley, 1962); idem, Nonlinear and Dynamic Programming (Reading, Mass.: Addison Wesley, 1964); Willard Zanguil, Nonlinear Programming: A United Approach (Englewood Cliffs, N.J.: Prentice Hall, 1969); and Michael Intriligator, Mathematical Optimization and Economic Theory (Englewood Cliffs, N.J.: Prentice Hall, 197 1).

4. Perhaps it bears pointing out that only final consumers use savings and loans in such a system. Production units do not borrow the "savings" of the "household sector" to "invest." The S&L's are simply clearinghouses for final consumers who wish to contract with one another concerning transfers of purchasing power through time.

5. Obviously, if wages affect work effort or if wage rates affect the educational and training preferences of individuals, which in turn affect the number of people who become qualified in different categories, the Central Planning Board cannot ignore efficiency effects when setting wage rates. But traditional welfare theory equates labor hired with labor done, and in Model 4 the availabilities of different kinds of labor in different time periods are givens. Alternatively, in some versions of central planning the supplies of different kinds of labor are "planned," just as the economy is planned, according to relative productivities and educational production coefficients. In either of these cases wages can be set without affecting the supplies of different kinds of labor.

6. This illustrates an interesting contrast between "individual" flexibilitythe freedom of individual personnel departments and employees to

negotiate-and "social" flexibility-the range of labor income distributions the system can generate. Of course, even under labor market system 3b, a system of one-time-only, positive and negative lump-sum taxes for all individuals in all periods that does not affect the "speed of the steeds" preserves the validity of the third fundamental welfare theorem for PuECPEs.

7. If the vi(t)'s and wi(t)'s in the social welfare function are determined independently of any prices and wage rates that may develop under any of the goods and labor market systems discussed previously, then those markets, prices, and wages have no effect on the production plan. That is, they have no effect on how much of each final good will be produced and how much of each kind of labor service will be used.

8. We mentioned in chapter 4 that Gintis proved Pareto optimality could be achieved in the general equilibrium of a particular type of economy even if people's preferences were endogenous. (Herb Gintis, "Alienation and Power: Towards a Radical Welfare Economics" [Ph.D. diss., Harvard University, May 1969], 282, Theorem 5). His model stipulated a production sector of the economy that adjusted efficiently to changing price signals for final goods and "primary" factors and a Central Planning Board that announced a particular equilibrium price path for final goods and primary factors that the production sector and individuals with endogenous preferences would react to. This model is very close to our own model of a centrally planned economy with free markets in labor services and final goods, where the weights for different arguments in the social welfare function are taken from their respective market prices. Rather than reproduce Gintis' proof under essentially similar assumptions, we refer the reader to his efforts for a demonstration that a centrally planned economy with free markets in labor and final goods can achieve optimality if individuals have perfect knowledge of their own endogenous preferences, even if the Central Planning Board does not.

9. E. Malinvaud, "Decentralized Procedures for Planning," in Activity Analysis in the Theory of Growth and Planning, E. Malinvaud and M. Bacharach, eds. (London: Macmillan, 1967), 170.

10. While dealing more "realistically" with information problems, it can be argued that all the procedures we review next are still "naive" regarding "incentive" problems. Specifically, all the procedures implicitly assume that local management responds truthfully to CPB queries. We discuss the "naivite" of this assumption later.

11. Benjamin Ward, The Socialist Economy: A Study of Organizational A Alternatives (New York: Random House, 1967), 45.

12. Ibid. If c is our column vector of final demands, or net output, the Central Planning Board will have succeeded in finding the vector of gross outputs, x, as: c + Ac + AAc + ... = x.

13. Ibid., 45-56.

14. J. M. Montias, "Planning with Material Balances in Soviet-Type Economies," American Economic Review 49 no. 5 (December 1959): 963-85.

15. There are ways to ameliorate the deficiencies of material balances Ward does not discuss. (1) Planners need not wait until the final iteration to check if the planned gross outputs are feasible. After each iteration resource feasibility could be checked, since there is no point in further iterations once feasibility is violated. (2) After the first response from all sectors the central planners can calculate the input-output matrix for the economy by dividing each sector's input requests by the final demand for the sector. If they asked each sector to respond with primary as well as intermediate input needs to meet the final demand, they could similarly calculate the economy's primary input coefficient matrix. If the central planners knew the social welfare function, they could then use the input-output matrix and primary input coefficient matrix to solve the "social optimization problem." This would allow substitution between different goods in the vector of final demands in light of productive possibilities. But this procedure would still be inefficient in the case of multiple processes. And it resembles some aspects of more efficient procedures we discuss below more than it resembles the traditional practice of material balances in any case.

16. George Danzig and Philip Wolfe, "The Decomposition Algorithm for Linear Programs," Econometrica 29, no. 4 (October 1961): 767-78.

17. Leonid Hurwicz, "The Design of Mechanisms for Resource Allocation," in Frontiers of Quantitative Economics, Vol. 2, Michael D. Intriligator and D. A. Kendrick, eds. (Amsterdam: North Holland Publishing, 1974),10.

18. Ibid.

19. William J. Baumol and Tibor Fabian, "Decomposition, Pricing for Decentralization and External Economies," Management Science I I (September 1964): 1-32.

20. Hurwicz, "Design of Mechanisms," 11.

2 1. Malinvaud, "Decentralized Procedures," 170-208.

22. Hurwicz, "Design of Mechanisms," 15.

23. Malinvaud, "Decentralized Procedures," 174-75. Malinvaud's "utility function," u(x) is our social welfare function. His "set of acceptable final consumption, X" is a feasibility constraint beyond what we posed in our model stipulating minimal quantities of food, shelter, clothing, etc. The rationale for such a feasibility constraint derives from requirements of work force survival.

24. Ibid., 177.

25. Ibid., 178.

26. J. Kornai and T. Liptak, "Two-Level Planning," Econometrica 33 no. I (January 1965): 141-69.

27. Hurwicz, "Design of Mechanisms," 11.

28. Ibid.

29. S. A. Marglin, "Information in Price and Command Systems of Planning," in Public Economics, J. Margolis and H. Guitton, eds. (London: Macmillan, 1969), 54-77.

30. Hurwicz, "Design of Mechanisms," 12.

3 1. G. M, Heal, "Planning Without Prices," Review of Economic Studies 36 (1969): 347-48.

32. J. H. Dreze and D. de la Vallee Poussin, "A Tatonnement Process for Guiding and Financing an Efficient Production of Public Goods," Core Discussion Papers No. 6922 (Belgium: Universite Catholique de Lourain, 1969).

33. E. Malinvaud, "The Theory of Planning for Individual and Collective Consumption," paper presented at the Symposium on the Problem of National Economy Modeling, Novosibirsk, 1970).

34. Hurwicz, "Design of Mechanisms," 17-18.

35. G. M. Heal, "Planning, Prices, and Increasing Returns," Review of Economic Studies 3 8 (1971): 28 1.

36. Hurwicz, "Design of Mechanisms," 12.

37. Heal, "Planning, Prices, and Increasing Returns," 28 1.

38. M. Aoki, "Two Planning Processes for an Economy with Production Externalities," Harvard Institute for Economic Research, discussion papers, no, 157 (Cambridge, Mass.: 1970).

39. Hurwicz, "Design of Mechanisms," 13.

40. T. C. Koopmans, "Analysis of Production as an Efficient Combination of Activities," in Activity Analysis of Production and Allocation, T. C. Koopmans, ed., Cowles Commission Monographs, no. 13, (New York 1951),3397.

41. Hurwicz, "Design of Mechanisms," 14.

42. P. A. Samuelson, "Market Mechanisms and Maximization," in Collected Scientific Papers of Paul A. Samuelson, vol. 1, J. E. Stiglitz, ed. (Cambridge, Mass.: NUT Press, 1966), 425-92.

43. K. J. Arrow and L. Hurwicz, "Decentralization and Computation in Resource Allocation," in Essays in Economics and Econometrics, R. W. Pfouts, ed. (Chapel Hill, N.C.: 1960), 34-104.

44. Hurwicz, "Design of Mechanisms," 14.

45. H. Uzawa, "Iterative Methods for Concave Programming," in Studies in Linear and Nonlinear Programming, K. Arrow, L. Hurwicz, and H. Uzawa, eds. (Stanford, Calif.: Stanford University Press, 1958), chap. 10.

46. Abram Bergson seems to be the first to have posed the relationship between central planners and plant managers as a principal-agent relationship. Abram Bergson, "Managerial Risks and Rewards in Public Enterprises," Journal of Comparative Economics 2, no. 3 (September 1978): 21125. But Bergson did not work out the implications of moral hazard in this setting. However, Pak-Wai Liu, Kent Osband, Pamela Brown, Jeffrey Miller, and James Thornton have recently gone a long way toward doing so. See PakWai Liu, "Moral Hazard and Incentives in a Decentralized Planning Environment," Journal of Comparative Economics 10, no. 2 (June 1986): 91-105; Kent Osband, "Speak Softly, but Carry a Big Stick: On Optimal Targets under Moral Hazard," Journal of Comparative Economics 11, no. 4 (December 1987): 584-95; P. C. Brown, J. Miller, and J. Thornton, "An Optimal Incentive Scheme for Planning with Targets," Journal of Comparative Economics 11, no. 4 (December 1987): 596-600; and Pak-Wai Liu, "Optimal Incentive Schemes with Targets: First Best or Second Best, " Journal of Comparative Economics 11, no. 4 (December 1987): 601-602.

47. See M. Keren, "On the Tautness of Plans, "Review of Economic Studies 39 (December 1972):469-86; M. L. Weitzman, "The New Soviet Incentive Model," Bell Journal of Economics 7, no. I (Spring 1976): 251-57; J. P. Bonin, "On the Design of Managerial Incentive Structures in a Decentralized Planning Environment," American Economic Review 66, no. 4 (September 1976): 682-87; J. Bonin and A. Marcus, "Information, Motivation, and Control in Decentralized Planning: The Case of Discretionary Managerial Behavior," Journal of Comparative Economics 3, no. 3 (September 1979): 235-53; M. L. Weitzman, "The 'Ratchet Principle' and Performance Incentives," Bell Journal of Economics 11, no. I (Spring 1980): 302-308; B. Holmstrom, "Design of Incentive Schemes and the New Soviet Incentive Model," European Economic Review 17, no. 2 (February 1982): 127-48; and David Granick, "Institutional Innovation and Economic Management: The Soviet Incentive System, 1921 to the Present," in Entrepreneurship in Imperial Russia and the Soviet Union, G. Guroff, ed. (Princeton, N.J.: Princeton University Press, 1983), 223-57.

48. Liu, "Moral Hazard," 91.

49. Malinvaud, "Decentralized Procedures," 17 1.

50. According to our work in chapter 7, public enterprise market economies have been as deserving as private enterprise market economies of such status, even though they have been accorded it less often. In any case, the important conclusion here is that public enterprise, centrally planned economies merit such status as well under the rules of the "old debate."

5 1. If laboring activities, and "bads" as well as "goods" are included in the social welfare function-as they should be-each person would assign a positive or negative sign to each argument before assigning whatever number of personal points he or she wished to that argument. The important thing is simply that the sum of the absolute values of all votes cast by an individual would be equal to the number of points he or she were permitted to vote.

52. Recent discussions and "reforms" in the Soviet Union and elsewhere have made this point more than academic.

53. Hurwicz, "Design Mechanisms," 38-39.

54. Ibid., 22.

55. Ibid., 16.

56. This includes all the gradient methods discussed earlier.

57. M. L. Weitzman, "Iterative Multi-level Planning with Production Targets," Econometrica 38, no. I (1970): 54.

5 8. Ward, Socialist Economy, 64.

59. Industry Ministries and provincial planning boards form intervening layers between production units and the CPB, but do not change the essential characteristic Ward identifies as the defining trait of hierarchic organizations: "no participant, directly or indirectly, is both superior to and subordinate to another."

60. Ward, Socialist Economy, 102.

61. While we may be less daunted than some by the formidable literature detailing conceivable difficulties with majority rule, we by no means deny these problems. For an excellent interpretation of the technical literature the interested reader should consult Alfred F. MacKay, Arrow's Theorem: The Paradox of Social Choice (New Haven: Yale University Press, 1980). But for now we waive all such objections because we wish to focus on a different problem with voting specific to central planning.

62. This last assumption might be more justified for a voting procedure in which everyone has the same number of points to vote and must use them to vote for public and private goods at the same time, than for market institutions. But "gamesmanship ... .. voting paradoxes," and "external effects" are admittedly difficult problems for any allocative mechanism to deal with successfully. While we allude to our belief that ideas from the literature on incentive-compatible mechanisms can be incorporated into a decentralized planning procedure that has more desirable properties than markets and central planning in our conclusion, we abstract from these problems here as irrelevant to our critique of central planning.

63. To argue that central planning is socially inefficient in this regard it is actually necessary to show that work in centrally planned economies,

regardless of occupation, will be characterized by less self-management than is possible under some alternative decision-making structure. The situation is completely analogous to the one we treated in chapter 8 regarding Bowles' claim that PrEMEs are inefficient because they aggravate malfeasance. The charge of social inefficiency always implies that there is some way of avoiding the problem-a feasible alternative that, at least, ameliorates the problem. We do not spell out the feasible alternative we prefer here, because in fact, demonstrating that our alternative model of participatory, decentralized planning is both feasible and optimal entails a complicated argument that we are still in the process of formalizing. But for now PuEEMMEs are sufficient as an alternative decision-making structure that provides work characterized by more self-management than PuECPEs. So the alternative of PuEEMMEs renders the charge that PuECPEs are inefficient in this regard complete.

64. If actual centrally planned economies provide no self-managed work opportunities, on any terms, for the vast majority of citizens, then the bias is even more extreme than we have assumed, and our case is all the stronger.

65. This would be the finding of traditional welfare theory if it were not encumbered by ideological biases, or if it were applied by the likes of a Leonid Hurwicz or Kenneth Arrow.

66. For our own thoughts on these subjects see Michael Albert and Robin Hahnel, "Ticket to Ride: New Locations on the Class Map" in Between Labor and Capital, Pat Walker, ed. (Boston: South End Press, 1979). Besides other essays in Walker's collection, and particularly the lead essay by John and Barbara Ehrenreich, see also Donald Stabile, Prophets of Order (Boston, South End Press, 1984), for an analysis of the development of the coordinator class in the U.S.

67. For full treatment see: Michael Albert and Robin Hahnel, Marxism and Socialist Theory (Boston: South End Press, 198 1); idem, Socialism Today and Tomorrow (Boston: South End Press, 1981); and idem, Political Economy of Participatory Economics, (Princeton, N.J.: Princeton University Press, forthcoming, 1990).

68. This has been the dominant view of international Trotskyism and Social Democracy over the years. See the writings of E. H. Carr, Alec Nove, and Isaac Deutscher for high-quality examples.

69. See the more recent writings of Paul Sweezy and Daniel Singer for highquality examples of this interpretation of the Soviet Union.

70. Howard Sherman, Foundations of Radical Political Economics (Armonk, N.Y.: M. E. Sharpe, 1987), 286-87.

7 1. See Albert and Hahnel, Socialism Today and Tomorrow.


1. As our analysis in part 3 indicates, "inappropriate assumptions" would be a more accurate qualification here. But regardless of their relevance, they are the assumptions typical of the traditional paradigm.

2. We are speaking here of properties of different theoretical models studied at the "highest" level of abstraction, that is, abstracting from information and incentive problems. As we discussed in chapter 9, the 11 modern" debate among traditional theorists is largely concerned with differences of this kind between economic systems. That is, the "modern" debate is conducted at a lower level of abstraction than the "old" debate. As we indicated, one reason for this shift might well be because at the highest theoretical levels traditional theory cannot meaningfully distinguish among any of the theoretical models studied.

3. We hasten to reiterate that under the usual assumptions not only will PrEMEs be (mis)perceived as containing no such biases, but PuEMEs and PuECPEs will be (mis)perceived as having no such biases as well.

4. We do not intend this as a criticism of Groves et al. We see no reason why societies heavily endowed with graduates of higher education should not enjoy the luxury of a division of intellectual labor that includes a role for those with a special talent for formal, abstract theorizing. We merely point out that there was no rush of economists with a more political "bent" to promote the theorists of incentive-compatible mechanisms for Nobel consideration!

5. For our technical formulation of a new economic model see Michael Albert and Robin Hahnel, Political Economy of Participatory Economics, (Princeton, N.J.: Princeton University Press, forthcoming, 1990). For a detailed account of the social relations of the new type economy we propose as it might look when implemented in the U.S., see Michael Albert and Robin Hahnel, Looking Forward: Participatory Economics in the 21st century (Boston, Mass.: South End Press, 1991).