![]()
1 Objective and scope of the work
Welfare economics is a systematic study of normative policy recommendations in economics. Applied welfare economics is an attempt to say “what ought to be” with respect to government policy, and theoretical welfare economics is the attempt to learn how to create a reliable applied welfare economics. The history of welfare economics is a strand of the history of the twentieth century. During the twentieth century, welfare economics emerged from neoclassical economic theory as a specialized field of study, only to strangely disappear from the mainstream of economics by the end of the century, as Sir Anthony Atkinson observed in 2001. Following mid-century, welfare economics was widely taught. (I was hired in my first faculty position in part because I was considered qualified to teach welfare economics.) Today it is far less commonly taught, and while some research continues, it is largely ignored. In Atkinson’s words (p. 193), “The ‘strangeness’ is that, despite the prevalence of welfare statements in modern economics, we are no longer subjecting them to critical analysis.” This book will attempt to trace and, perhaps, explain that rise and dissolution.
Now, economists in earlier centuries, like those of the twenty-first century, did not hesitate to make policy recommendations. Adam Smith could advocate for free trade and laissez-faire on the ground that it would increase the revenue of the nation; Ricardo could oppose the corn laws because they would raise the price of corn and shift income to the landlords. However, the emergence of neoclassical economics, particularly in the work of Marshall, made possible a narrower focus and more precise answers. Any excise tax will discourage production, no doubt, but there is no civilized society without some taxation. How much will a particular tax discourage production, and could it be that some taxes are less burdensome in this way than others? The apparatus of elasticity of supply and demand, consumer and producer surpluses showed the way to answer questions at that level of detail, questions that classical economics could not have answered. This is the opportunity that arose, for economists, with the maturation of neoclassical economics in the work of Marshall and Pigou. In Wealth and Welfare (1912) and Economics of Welfare (1920), A. C. Pigou presented a highly coherent account of the relation of economic activity to welfare as he conceived it, an account that was to influence both the construction of economic statistics that remains standard, and generations of research by economists.
There was, of course, some hesitation to adopt the Marshallian kit of tools. Many economists had followed Smith and Ricardo in arguing for free markets, for laissez-faire; others had argued the contrary. These large issues seemed to be the ones that mattered. To ask which taxes are less burdensome must have seemed a distraction from the more important question whether markets, on the whole, could better promote the prosperity of the many than could direct action by the state. There was probably less doubt about the value judgments that underlay the neoclassical tools. Advocates of free markets and advocates of state action could alike claim that the policies they advocated would encourage the prosperity of the many. The two sides in this great controversy shared the same, rather vague, value judgments. Their differences had to do with the nature of the economic system.
A further source of hesitation was the very precision of the Marshallian approach itself. To say that we can determine which taxes are less burdensome, and how much less burdensome, is to imply that the government ought to consider the taxes case by case, and choose those that are least burdensome. More generally, the government would be able to consider the merits of market processes as against regulation case by case. But, to the believers in laissez-faire, this would require the state to be the very meddlesome busybody they opposed. The precision of Marshallian analysis, if it could really be applied, was itself a defeat for their program. And the opponents of the free-market position were no more comfortable with the case-by-case approach. They were convinced that market systems were just chaotic and unreasonable, so that a simple set of bureaucratic rules could easily bring about a vast improvement of prosperity. Neoclassical economics, with a predictable logic of market equilibria and a focus on marginal improvements, must simply be wholly false, a fraud rather than a theory.
Further, the neoclassical apparatus was complex and tricky. This was a third major reason to hesitate: may we really be sure that these arcane graphs and computations could be a guide to the policies that would bring about greater prosperity of the many? What does consumer surplus really mean, and how should we really measure it? These questions would occupy many of the pages of the newly founded research journals in economics in the early years of the twentieth century, and several of Pigou’s early papers were important contributions.
The questions that remained largely unasked, even by those who hesitated to adopt the Marshallian approach, were those of value: just what constitutes the prosperity of the many, and why should it be a goal of public policy or of right action? Why should we favor policies that increase the revenue of the nation? Why should we not favor policies that shift income from the working class to the landlords? What, to use Pigou’s phrase, is the national dividend and why should we favor public policies that increase it? But these questions had to be addressed. The very precision of the neoclassical theoretical apparatus demanded that the value questions be addressed with similar precision. Pigou addressed them, and in the process reinforced a dilemma for many economists. Pigou’s welfare economics was firmly based in utilitarian ethical theory as Pigou understood it. This theory provided the most transparent support for the argument that free markets could result in increased prosperity in a morally significant sense. On the other hand, it had been known since the work of John Stuart Mill that utilitarianism could imply an argument for redistribution of income toward equality. Mill had also called for a separate consideration of efficiency and distribution. This was the dilemma economists had inherited from Mill, and to accept Pigou’s analysis of the efficiency of competition seemed to make a distinction between efficiency and distributive norms impossible.
Pigou also developed Marshall’s discussion of externalities, creating the modern concept of externalities. In the retrospect of a century, this may seem more important than the utilitarian case for (a limited amount of) redistribution. However, it was the Millian dichotomy of efficiency and distribution that motivated the attempt by the New Welfare Economics to reconstruct welfare economics on a different, though closely related, value theory. It will be argued that the creativity and success, and ultimate failure, of the New Welfare Economics led both to the rise and the disappearance of welfare economics in the twentieth century.
Pigou’s welfare economics is described as “highly coherent” in somewhat the sense of C. I. Lewis (1946), or Laurence BonJour (1985): a body of discourse is coherent only if it is logically consistent; and is relatively more coherent to the extent that it comprises propositions that entail or enhance the conditional probability of one another; and is less coherent to the extent that it can be divided into unconnected subsets of propositions and to the extent that it contains unexplained anomalies. This is not to assert a coherence theory of truth or justification, to assert, that is, that coherence is either necessary or sufficient to evaluate truth or to justify belief in all or part of the discourse. However, coherence is useful. In particular, it is useful to have coherence between the concept of economic welfare and the theory of the motivations, decisions and actions of people in the marketplace. If we were to adopt a concept of welfare and a theory of economic action that contradict one another or are without logical or probabilistic connections, we would raise the questions why (on the one hand) people are unmotivated by what we suppose to be their welfare or (on the other hand) whether a concept of welfare can be meaningful if it is unrelated to, or indeed contradicts, the motives that people reveal by their actions. Pigou’s welfare economics was highly coherent in that sense. A large part of the story to follow is of the attempt to retain that coherence without the specific utilitarian basis that Pigou gives it, and the failure of that attempt.
Pigou’s work did not of course take place in a vacuum, but quite the contrary, could reasonably be considered the maturity of the neoclassical tradition that had been growing in the last third of the nineteenth century. Much of this prehistory is well known and will not require a resume here. Some other parts – the utilitarian argument for redistribution, consumer and producer surpluses and the “measuring rod of money,” and externality – are best discussed along with the development of those ideas in Pigou’s writing and after. The ideas of utilitarians, and the incorporation of their ideas into economics after Mill and Jevons, deserve separate attention. It will be the topic of Chapter 2. Pigou’s welfare economics will then be discussed in Chapter 3. The fourth chapter will review the rise of the New Welfare Economics, its critique of Pigou and its attempt to refound welfare economics. Such further developments as the social welfare function, community indifference curves and the renewed interest in externalities will be considered in the chapters that follow. The challenges to welfare economics from the Arrow impossibility theorem, the second-best principle, and “the impossibility of a Paretian liberal” will be the topic of Chapter 7. Subsequent chapters will consider more recent developments, largely parallel but to a considerable extent independent, such as justice as fairness and fairness as the absence of “envy,” self-assessed happiness, behavioral anomalies and transaction costs. The last chapter is a brief summary and prospect.
Several ideas from twentieth-century philosophy will be called on. The work of Douglas Walton (1989, 2007, 2011) on informal logic and defeasible reason will play a double role. On the one hand, the book as a whole and its successive parts are organized around the judgment that the development of welfare economics at its best is an instance of what Walton calls a reasonable dialog.1 On the other hand, some scholars – Pigou, Little and Sen may be mentioned – have taken the view that discussions of public policy both among economists and in the population in general cannot avoid reliance on arguments that, while reasonable, are inconclusive. Here, again, Walton’s ideas and the concept of reasonable dialog are applicable. From the philosophy of perception and subjective experience, we will find the concepts of qualia and the associated ideas useful. These will be defined in Chapter 2 and used from time to time in subsequent chapters. For welfare economics, clearly, the relation of factual and valuative assertions is a key issue. Here, relatively recent philosophic discussions of “thick” concepts provide a standpoint for reconsidering economists’ debates about these matters. A thick concept is one that links an evaluative expression with one that is descriptive or non-evaluative, in such a way that the evaluation is merited because the object described is as it is in non-evaluative terms. Can we do welfare economics without relying on thick concepts? If not, what are the implications of this, and what may we learn about the economists’ controversies on this point in the twentieth century? Further discussion about thick concepts will be deferred to Chapter 5.
Note
1 Compare the retrospective view of Hicks (1975, p. 326), which concludes, “What I am saying, at the conclusion of this essay, is that Welfare Economics, in the broad as well as in the narrow sense, is itself a Critique.”
References
Atkinson, Anthony B. (2001) The Strange Disappearance of Welfare Economics, Kyklos v. 54, no. 23 (May–Aug) pp. 193–206.
BonJour, L. (1985) The Structure of Empirical Knowledge (Cambridge, Mass.: Harvard University Press).
Hicks, J. R. (1975) The Scope and Status of Welfare Economics, Oxford Economic Papers v. 27, no. 3 (Nov) pp. 307–326.
Lewis, C. I. (1946) An Analysis of Knowledge and Valuation (LaSalle: Open Court).
Pigou, A. C. (1912) Wealth and Welfare (London: Macmillan).
Pigou, A. C. (1920) Economics of Welfare (London: Macmillan).
Walton, Douglas (1989) Informal Logic: A Handbook for Critical Argumentation (Cambridge: Cambridge University Press).
Walton, Douglas (2007) Evaluating Practical Reasoning, Synthese v. 157, no. 2 (July) pp. 197–240.
Walton, Douglas (2011) Defeasible Reasoning and Informal Fallacies, Synthese v. 179, no. 3 (April) pp. 377–407.
![]()
2 Utilitarianism for Pigou’s welfare economics
The prehistory of the marginal utility theory of demand is well known and will be briefly sketched here for context. The classical political economists, from Smith through Mill and Cairnes, saw market prices as unrelated to utility. This was expressed in Smith’s paradox of diamonds and water. Accordingly, the classical political economists focused on a long-run, cost theory of price with labor as the measure of cost. There were, of course, some passages in the classical writings that pointed toward the marginalist theory. Malthus hinted, in a little-known essay (1800), that demand price would be the marginal willingness to pay, and his proposition of diminishing productivity of land would be generalized in neoclassical economics. John Stuart Mill had a modern conception of demand, applied in exceptional cases such as international trade and monopoly, though he did not derive it from utilitarian premises. Utilitarianism and price theory cohabited in Mill’s thought but he did not marry them, nor did they have issue.
Jevons, however, did marry utility and price theory, writing (1866, p. 282) “A true theory of economy can only be attained by going back to the great springs of human action – the feelings of pleasure and pain.” It is hard to imagine a more emphatic assertion of Benthamite utilitarianism, at least as a theory of individual economic choices. Yet one could – and many economists would – adopt Jevons’ theory of market equilibrium without adopting a Benthamite ethical theory or Benthamite norms of public policy. In itself, Jevons’ market equilibrium theory per se is not utilitarian. It is an explanatory theory. Jevons tells us that exchange will produce a surplus for at least one of those entering into exchange, but he adds by way of qualification that “There are motives nearly always present with us, arising from conscience, compassion, or from some moral or religious source.” However, as soon as Jevons’ successors and co-discoverers began to argue that the mutual gains from exchange justify a market economy and public policies that encourage a market economy, they were making a utilitarian judgment.
The ideas of Henry Sidgwick deserve some comment here, if only because it has been suggested that he anticipated Pigou’s foundation of welfare economics (Schultz 2017, p. 312). Sidgwick can be ranked along with Bentham and Mill among the great English nineteenth-century utilitarian philosophers (e.g. Schultz 2017, ch. 4). His book Methods of Ethics (1874) is probably the most important basis for this assessment. Sidgwick’s Principles of Political Economy (1883) is a restatement of the classical political economy incorporating, among other modifications, Jevons’ model of demand (pp. 72, 180). In Book III, “The Art of Political Economy,” Sidgwick addresses political economy as a public-policy discipline and relies mainly on utilitarian judgments as the decisive criteria. To this extent we may say that he raised the questions that Pigou would address. There is some difficulty of interpretation here. “As in the case of the Methods, the structure of these works has often made it difficult to see precisely where Sidgwick comes down on many issues” (Schulz 2017, p. 312). Sidgwick does say (1883, p. 518), like Pigou and perhaps less equivocally than Mill, that redistribution toward equality will ceteris paribus increase aggregate utility. In practice, though, Sidgwick never sees ceteris as paribus, systematically rejecting actual proposals for redistribution on acc...
