2. COMPARING ASSUMPTIONS Contemporary philosophy of science has had remarkably little impact on the theory or practice of economics. A simplistic and outdated brand of positivism has insulated the basic assumptions of neoclassical economic theory from critical scrutiny. Yet neoclassical theory, no less than Marxian theory, is largely structured by its basic assumptions. Kuhn argues that distinct "paradigms" are characterized by distinct, virtually incommensurable sets of assumptions. Whatever their other differences, however, both neo classical and Marxian economic theory seem to be wedded to a rosy picture of the household as home, sweet home
Economists often pride themselves on practising the most scientific of the social sciences. Yet as Thomas Kuhn and Thomas Feyerabend, among others, have argued, the borderline be tween science and other intellectual pursuits is difficult to draw (Kuhn, 1970 Feyerabend, 1975, 1978). Even the most "scientific" of theories are based on untestable or circular assumptions. They are seldom, if ever, verified or even falsified in any conclusive way by empirical research, Even more important. normal scientific research agendas are often limited to questions that can be answered simply by means of technical ingenuity. As Kuhn writes, "normal science does not aim at novelties of fact or theory, and, when successful, finds none" (Kuhn, 1970,
Both neoclassical and Marxian theory take somewhat circular assumptions as their starting points. The principle of utility maximization is "impregnable" because utility can be defined as whatever is being maximized (Meek, 1962). The principle serves primarily to generate very general predictions regarding the effects of changes in relative prices and incomes on individual behavior. The labor theory of value also holds true by definition. It merely asserts that value can be defined as the amount of socially necessary labor time embodied in a good. and that value bears a determinate mathematical relationship to prices (Steedman. 1977). Individual utility func- t ions are unobservable, and because revealed preference can be ranked only in ordinal terms. they cannot be aggregated or compared (Arrow. 1963). By the same token, socially necessary generations of labor time (which includes i indirect labor inputs) cannot be directly calculated, in part because it requires the aggregation.
To take another pair of examples, neoclassical theorists assume that individuals (or household that behave as though they were individuals) and firms are the primary actors in an economy, and they interact almost exclusively through competitive markets in which no single individual ot find can affect supply and demand. They cling to methodologies based on this assumption despite considerable evidence that few if any markets are perfectly competitive (Friedman, 1953). Marxian theorists tend to use classes as their unit of analysis, defining classes in terms of relationship to means of production and control over the labor process. They pursue the methodology of class analysis despite the evidence of more diverse and complicated forms of social stratification (Poulantzas, 1975; Wright, 1979).
Such assumptions, however unrealistic. seem to be an indispensable means of structuring research. Acknowledging this unpleasant fact, positivists argue that assumptions themselves merit little scrutiny, that theories should be judged solely by their success at generating and satisfying empirical predictions. Milton Friedman's classic statement of this view in "The Methodology of Positive Economics," continues to be widely quoted by economists, who tend to respond to criticisms of their assumptions simply by reiterating their empirical results. This response is not particularly convincing when. as Thomas Kuhn and others have pointed out. the same empirical results may be consistent with more than one theoretical construction (Kuhn, 1970,
The lack of realism inherent in most theoretical assumptions does not. however, render com- between them inconsequential or unnecessary. Among the criteria that are often used to compare alternative sets of theoretical assumptions are internal logic. consistency. and range of explanation. Marxian economics. largely excluded from the "dominant" paradigm, not positive or empiricist views hut occasionally tend to the opposite extreme. Barry Hindess and Paul Hirst (1975), for instance question the relevance of historical "facts" and elevate the role of deductive logic. Steven Resnick and Richard Wolff (1982) argue that neither assumptions nor facts provide adequate criteria for choosing between theories. Different theories merely represent different "points of entry." chosen for historical or personal reasons.
impossibility of specifying any absolute criteria for success or failure does not vitiate the need for constant evaluation. As E. P. Thompson writes, the evidence "is there, in its primary form, not to disclose its own meaning, but to be interrogated by minds trained in a discipline of attentive Thompson, 1978, p. 28). Such attentive disbelief can be enhanced by efforts to compare contending theories and "translate" them into commensurable terms. Such an analysis must go well beyond a comparison of competing hypotheses. As Foucault (1973) observes. the silences within theories often speak more loudly than the claims Why are both the neoclassical and the Marxian paradigms so "silent" on the issue of inequality within the home? Their convergence is somewhat tonic: On the other hand we have a paradigm, largely unconcerned with issues of conflict. which offers a well-developed theory of nonmarket induction. On the other hand, we have a paradienn with a well developed theory of conduct that is largely unconcern cd with nonmarket prediction.
As historians of economic thought have note neoclassical theory offers a "universal" theory of economic that presumes the existence but is of any institutional context (Hunt, 1979). Markets are definition. sites of free and attention to change. Any theory that restricts its an explana the market will be unable to provide or In a exploitation, in the economic sense. can nest parties benefit from a transaction it will not take place. The or inequality of any given exchange cannot be ascertained because interpersonal comparisons cannot be made
The plausibility of the "new home economies tests on an analogy between the household and the firm, where individuals within the household operate in an implicit" rather than explicit market. Gary Becker (1976) theorizes, for instance, that new households are formed by means of a marriage market. Once established, households deploy labor in response to differ- ences in marginal productivity between home and market (Gronau, 1973). They choose be tween home produced and market produced goods on the basis of price, although some goods, like one's own children, can only be produced at home. Decisions about family size are influenced by changes in the price of children. due to increases in production costs such as education (sometimes termed improvement in child quality) or the increased opportunity cost of time devoted to childcare (Schultz. 1981)
As many critics have noted, the analogy between the household and the firm is tomewhat trained. Many different types of exchange take place within and among household but these exchanges do not conform closely to sales and purchases in a competitive market. Individuals seeking mates are usually not satisfied with an "undifferentiated" product. In household production, unlike commodity production, demander and suppliers are often one and the same. The cost of home produced goods, like children, is not exogenously determined. Parents choose, to some extent, how much time and how many resources to devote to children. And children affect their own price. As adults they make certain choices concerning their economic commitments to aging parents (Parsons, 1983) Finally and perhaps most importantly, the neoclassical household does not have an objective function that is as clear-cut as that of the neoclassical firm. It does not maximize profits. It presumably maximizes the "joint utility" of its et members
The specification of joint utility function poses a serious problem for neoclassical theory because it requires the aggregation of individual tastes and preference, a task that is intrinsically problematic (Arrow. 1963). One way of solving this aggregation problem is to assume that altruism prevails within the family (Samuelson, 1956). But it is somewhat inconsistent to suggest that individuals who are entirely selfish in the market (where there are no interdependent utilities) are entirely "selfless" within the family, where they of pure altruism within the family resembles nothing so much as the Marxian vision of utopian socialism. There is something paradoxical about the juxtaposition of naked self-interest that presumably motivates efficient allocation of mar-? ket resources and a perfect altruism that presumably motivates equitable allocation of family resources
Lest Gary Becker unfairly be assigned responsibility for this paradox. it should be noted that it has distinguished roots in the history of economic thought. One of the most widely quoted passages of Adam Smith's Wealth and Nations observes, "It is not from the benevolence of the butcher, the brewer, or the baker that we expect our dinner but from their regard to their self-interest" (Smith, 1937, p. 14). But as Smith made clear in The Theory or Moral Sentiments. (1966), he was not so skeptical of the benevolence of fathers and husbands. Becker (1976, 1981) seeks to provide scientific basis for his modern explanation of "moral sentiments," drawing from the sociobiology