1. Economic methodology
Philosophical reflection on economics is ancient, but the conception of 'the economy' as a distinct object of study dates back only to the 18th century. Aristotle addresses some problems of economics mainly as problems of household management. Scholastic philosophers addressed ethical questions concerning economic behavior, and they condemned 'usury' -- that is, the taking of interest on money. With the increasing importance of trade and of nation-states in the early modern period, 'mercantilist' philosophers and pamphleteers addressed questions concerning the balance of trade and the regulation of the currency. Only in the work of the physiocrats and especially of Adam Smith do scholars begin to think of the economy as an object of study with its own principles and laws.
18th-century philosophers wrote in the shadow of Newton's towering accomplishments. David Hume is unabashed about his hopes to develop a science of mind and society in the image of Newton's science of the solar system. To that end, he seeks out general laws of individual thought and action, out of which larger-scale orderly relations will arise, in just the way that patterns in planetary motion arise from the laws of motion and gravitation governing individual bodies. Thus Hume traces the rise in prices and the temporary increase in economic activity that follow an increase in currency to the perceptions and actions of individuals who first spend the additional currency.
It then lay to Adam Smith to extend such a method to a systematic Inquiry into the Nature and Causes of the Wealth of Nations and to make explicit an implication of previous causal inquiries such as those of Hume, the physiocrats and of many of the mercantilists. That implication (which is arguably the foundation for all social science) is that the social, aggregative implications of individual choices are often unintended. In spending their additional gold imported from abroad, traders do not intend to increase the price level. But that is what they do nevertheless. In this way unintended regularities that arise from the intentional choices of individuals can rule those individuals just as surely as the regularities of nature, and, like nature, societies can be objects of scientific investigation.
|Adam Smith, the father of Modern Economics|
1.1 Classical economics and the method a priori
Although Adam Smith made some general comments about scientific method in his History of Astronomy and other essays, he wrote little explicitly about the method of economics. The first extended reflections on economic methodology had to wait until Nassau Senior (1836) and John Stuart Mill (1836). Their essays must be understood against the background of the prevailing economic theory. Like Smith's economics (to which it owed a great deal) and modern economics, the 'classical' economics of the middle decades of the 19th century traced economic regularities to the choices of individuals facing social and natural constraints. But, as compared to Smith, more reliance was placed on severely simplified models. In David Ricardo's Principles of Political Economy, a portrait is drawn in which wages above the subsistence level lead to an expanding labor force, which in turn requires more intensive agriculture or cultivation of inferior land. The extension of cultivation leads to lower profits and higher rents; and the whole tale of economic development leads to a gloomy stationary state in which profits are too low to command any net investment, wages return to subsistence levels, and only the landlords are affluent. From the time Ricardo's Principles was published (1819) the data available to classical economists was never in accordance with the trends the theory predicted. Yet the theory continued to hold sway for more than half a century, and the unfavorable data were explained away as due to various 'disturbing causes.'
It is not surprising then that Mill's account of the method of economics would emphasize the relative autonomy of theory. Mill distinguishes between two main kinds of inductive methods. The method a posteriori is a method of direct experience. It is only suitable to phenomena in which few causal factors are operating or in which experimental controls are possible. Mill's famous methods of induction are detailed specifications of the method a posteriori. In his method of difference, for example, one holds fixed every causal factor except one and checks to see whether the effect ceases to obtain when that one factor is removed.Unfortunately, such direct inductive methods cannot be used to study phenomena in which many causal factors are in play. If, for example, one attempts to investigate whether tariffs enhance or impede prosperity by comparing the prosperity with and without tariffs, the results will be irregular and unreliable because other causes besides tariffs will also differ across societies and times. So one needs instead to employ the method a priori. Despite its name, this is an inductive method, but it is an indirect inductive method. One first determines the laws governing individual causal factors in domains in which direct inductive methods are applicable. Having then determined the laws of the individual causes, one investigates their combined consequences deductively. Finally, there is a role for 'verification' of the combined consequences, but owing to the causal complications, this testing has comparatively little weight. The testing of the conclusions serves only as a check on one's deductions and as a indicator of whether there are significant disturbing causes that one has not yet accounted for. Mill gives the example of the science of the tides. One determines the law of gravitation by studying planetary motion, in which gravity is the only significant causal factor. Then one develops the theory of tides deductively from that law and information concerning the positions and motions of the moon and sun. The implications of the theory will be inexact and sometimes badly mistaken, because many subsidiary causal factors influence tides. By testing the theory one can uncover mistakes in one's deductions and evidence concerning the role of the subsidiary factors. Because of the causal complexity such testing does little or nothing to confirm or disconfirm the law of gravitation, which has already been established.
Because economic theory includes only the most important causes and necessarily ignores many minor causes, its claims, like claims concerning tides, are inexact. Its predictions will be imprecise, and sometimes dead wrong. But it is possible nevertheless to develop and confirm economic theory by first establishing in simpler domains the laws governing the major causal factors and then deducing their consequences in different circumstances. For example, statistical data tell a mixed story about the relationship between minimum wages and unemployment; and there is no data at all about what the consequences for employment would be of an extremely high minimum wage. On the other hand, everyday experience teaches one that firms can choose among more or less labor-intensive processes and that a high minimum wage will make more labor-intensive processes more expensive. Since one also has good reason to believe that firms try to keep their costs down, one has good reason to believe that a high minimum wage will increase unemployment.
According to Mill, economics is not only inexact and committed to the method a priori. In addition, he maintains that it is a separate science. What distinguishes economics as a discipline is not only its concern for a certain domain of phenomena, but also its limitation to a particular set of causal factors, which predominate in this domain. With respect to this domain, one can (at a certain level of approximation) ignore the myriad causal factors that influence all social phenomena and which are the subject matter for sociology and develop economics separately. In defending a view of economics as in this way inexact and separate and a view of economists as following the method a priori, Mill was able to reconcile his empiricism and his commitment to Ricardo's economics.
Although Mill's views on economic methodology were challenged later in the 19th century by dissident economists who believed that the theory was too remote from the contingencies of policy and history, Mill's methodological views dominated the mainstream of economic theory for well over a century. Mill's vision survived the transformation of economics from classical to neoclassical and is clearly discernable in the most important methodological treatises concerning neoclassical economics, such as John Neville Keynes' The Scope and Method of Political Economy (1891) or Lionel Robbins' An Essay on the Nature and Significance of Economic Science (1935). Indeed Hausman (1992) argues that current methodological practice closely resembles Mill's methodology.
|David Ricardo, one of the greatest Classical and Political Economist|
|John Stuart Mill|
1.2 20th-century philosophy of science and economic method
Beginning in the 1930s, mainstream economists began to have a bad conscience about their traditional methodology, which some saw as insufficiently empiricist. Robbins is a transitional figure, because at the same time that he clung to Mill's methodology and to the identification of the domain of economics with the predominance of particular causal factors (the allocation of scarce means that have alternative uses , he also offered his apparently empiricist critique of interpersonal utility comparisons as untestable value judgments . Terence Hutchison argued that the propositions of pure theory were so hedged with ceteris paribus conditions that they were not testable . Paul Samuelson argued for a need to separate the 'operationally significant' economic wheat from the chaff, and in his theory of revealed preference he provided a model for how to do so. Other economists cited survey data to argue that the theoretical propositions of economics were false. The confusing methodological situation stabilized in the 1950s with arguments by Fritz Machlup and especially Milton Friedman that economists need be concerned to fit only data concerning prices and quantities and that the 'realism of assumptions' is irrelevant.
Although confused, mistaken, and inconsistent with the practice of mainstream economists-including his own practice--Friedman's views have dominated the methodological conceptions of mainstream economists over the past two generations. His views are confused because they conflate many different kinds of 'assumptions' and 'realism'. Basic generalizations, propositions concerning initial conditions, and antecedents of conditional claims are all called 'assumptions'. Assumptions are called 'unrealistic' if they are false, incomplete, or not approximately true. Friedman's views are mistaken, because (as Mill already emphasized) market data are generated by too many causal factors to provide effective tests of economic theory. And the eclecticism that Friedman's official methodological position counsels--'don't worry about what theory says, just ask whether it fits market data'--contradicts the firm and narrow commitment to a specific theory that characterizes mainstream economics, including Friedman's own work. Nevertheless Friedman's views were warmly embraced because they freed mainstream economists to ignore criticisms of their theories and to abandon all empirical work apart from econometrics.
In the 1970s and 1980s other streams of contemporary philosophy of science began to influence the methodological reflections of economists. Karl Popper's view that scientific theories must be falsifiable, which had been defended by Hutchison was taken up and emphasized by Mark Blaug . According to Popper, the hallmark of science is to formulate theories so that they can be exposed to empirical testing and to reject theories that fail tests. Hutchison and Blaug object that mainstream economic theories are not harshly tested and that mainstream economists are unwilling to surrender them when they fail the few tests they are subjected to. In his extensive publications, Lawrence Boland has also emphasized the relevance of Popper's views to economic methodology, although Boland's interpretation of Popper is rather different than Blaug's or Hutchison's.
Imre Lakatos' methodology of scientific research programmes (which represented a wedding of some of Thomas Kuhn's insights to Popper's philosophy of science) was for a while widely discussed . Although still emphasizing the importance of empirical criticism, Lakatos insisted that theories should not be abandoned until superior alternatives are found, and his emphasis on heuristics struck a responsive chord.
A spate of prominent contemporary writers on methodology: Roger Backhouse, Bruce Caldwell, Neil deMarchi, D. Wade Hands, and E. Roy Weintraub have a more ambivalent relationship to Popper's and Lakatos' philosophy. Caldwell wrote a series of searching, though charitable criticisms of Popperian views. Backhouse, DeMarchi, Hands, and Weintraub were all at one time enthusiastic about Lakatos' views, though to differing degrees all have moved away from them now. Backhouse and deMarchi remain the closest. Hands has become a proponent of the application to economics of new work on the sociology of science, while Weintraub's work on the stabilization of concepts in economics has been more influenced by literary theorists.
More recently, alternatives to philosophy of science influenced by work in literary theory and sociology have also developed a following among economists. In a series of stylistically brilliant works, Deirdre McCloskey has criticized the whole project of exploring a normative methodology for economics and has urged instead that economists attend to their rhetoric--that is, to their ways of persuading one another . This work has been highly controversial, because many of McCloskey's formulations apparently reduce questions about the correctness or incorrectness of economic claims to questions about what most economists accept. Such a view would imply that minority views are always mistaken. McCloskey denies such a radical reading of her position, but it is not clear how to avoid it without incorporating normative methodological commitments into the rhetoric of economics.
Philip Mirowski's work on methodology is heavily historical and less radical epistemologically than McCloskey's. He has explored the influence on economics of the formal analogy between utility theory and physics (1990). More recently, his work has shown increasing concern with sociological influences. For other work that tests the boundaries of conventional methodology.
Over the last fifteen years, economic methodology has become a large field. It has its own Journal of Economic Methodology and occupies a large part of the journal, Economics and Philosophy. Dozens of monographs on economic methodology have been published. Graduate programs in the field have been established (including a Ph.D. program at Erasmus University in Rotterdam), and there are regular sessions on economic methodology at meetings of economists and philosophers. The field is very diverse, and each of its leading figures has a distinctive approach. Although one can roughly identify Popperian-Lakatosian and sociological-literary schools, there is no easy way to categorize contemporary methodology.
Several methodologists have been particularly interested in the role of explicitly causal notions in economics (see Causes and laws; Causation: physical, mental and social). Nancy Cartwright (1989), who is also a distinguished philosopher of physics, has (unlike most philosophers writing on economics) paid a great deal of attention to econometrics. She has defended the importance of specifically causal considerations in science, and she has argued that econometricians have made important contributions to a philosophical understanding of causation. Cartwright holds that economists should be understood as attempting to identify causal capacities and that the work of econometricians can be understood as contributing to this task.
Daniel Hausman (1992) defends a modernized variant of Mill's methodology that depicts economics as inexact and permits the plausibility of the basic principles of economics to count in the assessment of economic theories. But he is skeptical of the view that a small set of causal factors do indeed predominate in the domain of economics and defends a greater attention to the results of experiments, surveys, and field work. He has increasingly emphasized the role of causal generalizations in mainstream microeconomic theory.
Kevin Hoover (2000)-- unlike most methodologists -- has written on problems concerning macroeconomics. Although he has addressed technical issues in econometrics, his main interest has been in causality. Like Cartwright, Hoover has placed particular emphasis of what one can learn about causation by studying the work of economists.
Tony Lawson (1997) finds his inspiration in the 'transcendental realism' of Roy Bhaskar. Lawson sees himself not merely as enunciating a distinctive approach to methodology, but as laying the foundations for a new economics. Crucial to the approach and to his methodological views is a strongly realist ontology (see Realism, instrumentalism, fictionalism), which takes the objects of scientific investigation to be causal mechanisms and tendencies that lie as it were 'beneath' the irregularity of phenomena. This realist ontology leads him, like Cartwright, Hausman, to regard economic theories as identifying causal mechanisms and tendencies.
There is a great deal of other work, too. A number of economists and philosophers have attempted to apply to economics the structuralist view of scientific theories developed by Patrick Suppes, Joseph Sneed, and Wolfgang Stegmller. Alexander Rosenberg, who is also a distinguished philosopher of biology, has pursued a number of different themes. His Microeconomic Laws: A Philosophical Analysis (1976) was one of the first philosophical treatments of the methodological problems of economics. In that book Rosenberg argues that economics fits reasonably well within a standard philosophical model of the natural sciences. But shortly afterwards Rosenberg changed his views radically, and he now defends the position that mainstream economics is best understood either as applied mathematics or as a normative social and political theory (1992).
Contemporary economic methodology thus moves in many directions. Although some of it is directed toward the conduct of economics, one regularly finds a sort of methodological schizophrenia whereby in theory economists cling to positivist or Popperian philosophical views that are radically at odds with their practice, which is roughly Millian. Other work in methodology is oriented toward philosophy of science, and it appears that the study of economic methodology has made significant contributions to philosophical investigations of causation and explanation. It is controversial whether methodology can or should be a distinct and relatively self-contained field rather than addressing itself toward economists or toward philosophers.
2. Methodology, rationality, preference, and self-interest
The study of rationality, preference, and self-interest is a second area of overlap between economics and philosophy. Mainstream economic theory is built around a variant of 'folk psychology'. According to folk psychology, human actions are the consequence of beliefs, desires, and circumstances that determine the feasibility and consequences of actions. The same elements figure in economic theories of choice. Sometimes economists suppose that agents have complete knowledge and thereby avoid having to refer specifically to beliefs, but in other contexts beliefs are explicitly modeled--most typically as subjective probability judgments. In place of desires, economists postulate that economic agents possess stable, complete, and transitive preferences. Given additional technical conditions, such preferences can be represented by a continuous 'utility function' such that UP(A) > UP(B) if and only if an agent P prefers A to B. 'Utility' is simply a way of indicating how an alternative is ranked by an agent. It is not a substantive object of choice, and indeed it makes no sense to regard an agent as seeking or preferring utility. Mainstream economic models relate choice to preference by maintaining that, subject to the constraints, agents maximize utility -- which means merely that among the feasible alternatives agents choose whatever they most prefer.
The previous paragraph presents the standard theory of choice as an empirical theory purporting to describe, predict, and explain how agents choose in fact. But the same theory also functions as a theory of rationality: Preferences are rational if they are complete and transitive (and perhaps satisfy other axioms, too). Degrees of belief are rational if they satisfy the axioms of the calculus of probability. Choices are rational if they are utility maximizing. Given this model of rationality, the empirical claim of the standard theory of rational choice can be restated simply as the claim that economic agents are in fact rational.
According to this theory of rationality, it is irrational to choose one feasible option over another when one prefers the second and according to this theory of choice people will in fact never choose one feasible option over another if they prefer the second. Since according to these theories choice is and ought (rationally) to be determined by the agent's preferences, it is easy mistakenly to infer that these theories maintain that agents are and ought to be self-interested. But whether a choice is self-interested depends on the content of the preferences that lead to it, not on whether the choice is determined by preferences. People who prefer to sacrifice their interests to others are not self-interested, and that fact that their choices result from these preferences does not make their choices self-interested. Unlike the standard theories of choice and rationality, the views that it is rational to pursue one's own interest and that people in fact do so are 'substantive' theories of choice and rationality. Unlike the standard 'formal' theory, which merely specifies a certain structure of choice and preference, self-interest theories specify an objective that agents do or ought rationally to have.
Theories of rationality are normative theories. They prescribe what one ought to do--not, to be sure, as a matter of morality, but as a matter of rationality or prudence. (Irrationality is foolish rather than evil.) Incorporating as it does a normative theory, economics is unlike any of the natural sciences. The reason why economics incorporates a theory of rationality is that human actions, unlike the actions of oak trees or potassium, can be criticized or justified as well as explained. This fact has important methodological consequences and indeed (as just argued) establishes one important distinction between the social and the natural sciences.
The first and perhaps most important methodological consequence is that explanations of individual choices cite the reasons why the agent acted. Although a number of philosophers in the 1950s, who were influenced by Wittgenstein, argued that such explanations that cite an agent's reasons could not be causal explanations, nowadays most philosophers are persuaded by Donald Davidson's argument (1963) that satisfactory explanations that cite an agents reasons must also be causal explanations. Agents may reasons for performing an agent that are not in fact responsible for the action. Davidson argued that the distinction between the reasons that are 'effective' and those that are merely rationalizations is that the first unlike the second are causes of the action.
So the fact that the explanations economists offer of individual choices cite reasons for those choices does not imply that economists are not giving causal explanations, and it does not imply a strong anti-naturalist distinction between the structure of the natural and social sciences. But the fact that the causal explanations economists offer also cite reasons is nevertheless of great interest. It explains why economists must hold that agents are to some extent rational. (If they weren't then they couldn't have reasons for their actions.) Furthermore, it implies that the actions of economic agents are subject to rational appraisal. Were the agent's reasons good reasons? Was the action justified? And once one begins appraising choices, one is just one step away from ethical questions. Readers should be aware that the standard theories of rationality and choice are controversial.
3. Welfare, justice, equality, and freedom
Economists have an ambivalent attitude toward ethics. On the one hand, many have been concerned to insist that mainstream economics is a 'positive' science, whose conclusions are entirely independent of any moral commitments. On the other hand, economists freely give out normative policy advice. Some of this advice is purely technical, like the advice that a civil engineer might offer on where to locate a bridge, but much of it is not. Most mainstream economists are in fact committed to a specific view of ethics that emphasizes welfare, and they also adopt a distinctive theory of welfare. (When economists emphasize efficiency they are virtually always concerned with efficiency at promoting individual welfare.) These features of normative economics are not entailed by positive economics or the theory of rationality, but they are heavily influenced by them.
There are a variety of different ethical bases for assessing social arrangements. In addition to considering the welfare of the members, one can also ask whether the rights of the members are protected and whether procedures and the distribution are just, whether the members are treated equally, and what sorts of freedoms and opportunities people enjoy. All of these considerations are important, but in economic policy pronouncements and normative theory (which indeed is called 'welfare economics typically only considerations of welfare enter.
The reason for this narrow normative focus is that welfare can be tied tightly to the standard models of rationality and choice. Suppose that individuals are rational and--in addition--that they are self-interested. As already pointed out, the standard theory of rationality does not imply self-interest, but self-interest is also commonly assumed in most economic models. It follows that individuals will prefer alternative A to B if and only if they believe that A serves their interests better than B. If individuals have perfect knowledge, which is again a common assumption in economic models, it follows that individuals will prefer A to B if and only if A in fact serves their interests better. Finally, if one identifies an individual's welfare with whatever is in the individual's interest, it follows that A is better for and individual than B if and only if the individual prefers A to B--or in other words that welfare is the satisfaction of preference. This view of welfare has the additional advantage that it prevents concerns about paternalism--to which many economists are opposed--from even arising, since by definition it could never be good for individuals to overrule their preferences.
Most normative economics is thus doubly narrow. Not only is it characterized by an almost exclusive preoccupation with welfare, but it is also committed to a particular view of welfare as the satisfaction of preferences. This double narrowness is doubly unfortunate. The limitation to considerations of welfare means that in policy discussions economists treated non-welfarist ethical criteria as exogenous constraints to be left for someone else to worry about. The commitment to a preference satisfaction view of welfare has the drawback that this view of welfare is false. Because individuals are not always self-interested and because their beliefs are not always true, individuals do not always prefer what is good for them. There are a variety of good reasons to be hesitant about paternalism, but the view that individuals are always perfect judges of their own good is not one of them. There is some work, particularly by Amartya Sen , that stretches these limits on normative economics, but such work is the exception to the welfarist rule
Mainstream economists, whether engaged with positive theory, the theory of rationality, or with normative and policy investigations, are overwhelmingly committed to an exaggerated simplification of a plausible view of individual agents as possessing stable and consistent preference rankings, which, given the constraints, determine their choices. Market outcomes are the unintended consequences of these choices. Rationality is defined by this structure of choice. Welfare, like choice, is determined by these same preference rankings. Although there are philosophical issues concerning economics that have no particular connection to this core commitment, most work in philosophy of economics has attempted to comprehend and evaluate the basic model of choice and its applications in positive economics, normative economics, and the theory of rationality.