Dangers of Formalizing and Testing Economic Theory

Dangers of Formalizing and Testing Economic Theory

Ekelund and Hébert (1990), among other points, argue that over-formalization of economics using mathematical and statistical model building without the concomitant guiding vision of a problem to solve will be detrimental to the future of economic study. Are our present day scholars of economics well-equipped with tools, but bankrupt of original economic problems to research? It’s worth examining each of these major points in detail.

Ekelund and Hébert (1990) draw several conclusions about the future of economics. First, econometrics is emphasized in graduate study and its mastery is required for advancement in the profession. Second, econometrics is so deeply prevalent that further use should be weighed against its expediency and costs. Third, proponents of the advancement of the scientific nature of economics have not fully considered whether it is possible to achieve that goal. Fourth, those who question continued formalization rightly argue that economic theory is not totally verifiable because of the ethereal nature of human behavior and the extremely high costs of collecting relevant, associated data. Fifth, economic ideas have historically led or paralleled the development of the techniques to test them, rarely vice versa, and to reverse the order of this discovery process runs the risk of dividing modern economics into two disciplines: a branch of mathematics and a branch of the social sciences. Finally, it is not the blind analysis of data in the pursuit of an elusive truth that makes economics relevant, but the healthy discourse among those of varied points of view.

Ekelund and Hébert (1990) point out that the study of mathematics and econometrics form the basis of graduate economic curricula. University hiring policies and professional advancement within the discipline all require mastery and application of mathematical and empirical technique. Because this shift can be traced to before the second half of the 20th century, the authors’ view is that it heralds the future of direction of economics as being toward mathematics and away from the social sciences.

Is graduate training a rut from which no scholar can lift himself? Whether this assertion is true can be debated. Every great economist in history went far beyond his formal training, often breaking with tradition in arduous academic journeys. Clearly those who are trained in mathematical analysis will have a natural tendency to rely on those techniques, but to suggest that they will not employ verbal or graphical descriptions of relevant economic phenomenon is far-fetched. Smith, Ricardo, Mill, Marshall, Jevons, Walras, Veblen, and Keynes all pioneered new patterns of thinking and made original contributions to economic thought. Moreover, many of them did not even have the benefit of formal training in economics (often because the discipline did not formally exist.) So if the history of economics is any predictor of the future, one cannot prove that econometric digressions in graduate training alone will forever doom the discipline.

Ekelund and Hébert (1990) assert that econometrics and mathematics have permeated modern and microeconomic and macroeconomic theory and that deeper use of econometrics in the discipline of economics should be evaluated based on weighing the costs, the advantages and disadvantages. From the vantage point of scholarship, an analysis of the advantages and disadvantages of perpetuating mathematical techniques may be difficult to ascertain. Clearly, to the degree that it is possible to assess the future usefulness of present development of econometric technique, such surveying should take place. However, the historian of economic thought has the unique perspective of being able to see past digressions in the discipline and employ 20-20 hindsight on those events. My view is that the permeation of microeconomics and macroeconomics with mathematical techniques is not a cause for alarm as long as progress in theory and policy is being made.

Advocates of continued formalization of economic theory through mathematics argue that theories must be formalized and verified if economics as a discipline will achieve the completeness and be afforded the respect of a science. Ekelund and Hébert (1990) make the salient point that while this may be a necessary or an even important goal, its pursuit seems to have taken precedence over whether it is an achievable goal. The central theme in this portion of the debate is that economists do not get the respect they really deserve. Economic theories that have not been proven do not carry the weight of gospel and are difficult to translate into economic policy.

Should economics assume the rigidity of a formal science? This is an important question with far-reaching implications. Would it not be better to view it as Ekelund and Hébert (1990, p. 604) suggest, “a powerful, though somewhat imprecise behavioral science?” Even truths we hold dear in the physical sciences are not absolute nor without founding assumptions, on which the entire discipline rests. With this in mind, economic policy must be formulated whether or not economics is regarded as a precise and formal science. Imagine the result if the political economists simply had no recommendations for improving the post-mercantilistic economies because of the difficulty in collecting data or the incomplete testing of theories. Absurd!

Opponents of continued mathematization and formalization argued that economics is essentially a social science, and as such is subordinate to the whims of human behavior and subject to the inexactness and vagaries of small data samples and incomplete theories describing economic phenomenon. Furthermore, the authors of the text indicate that the critics of econometrics further argue that data in the necessary quantities and qualities to validate the economic theories in question can be obtained only at a prohibitive cost.

While it may be true that economics is essentially a social science, all great strides of progress in the past study of economics have had both theoretical and methodological components. In other words, the development of methodology to test theory was closely associated or developed in parallel to theory. For example, it was economists such as Edgeworth, et al, who developed descriptive statistical techniques powerful enough to serve as the foundation for modern statistics and econometrics. Cournot was one of the first to recognize the importance of mathematical tools in accurately encapsulating economic ideas to avoid the digressions into vague argumentation that could occur when economic analysis was explained only in literary form. One possible danger with widespread use of econometrics is the sheer size of the data collection for analysis – economic problems that do not lend themselves to this sort of analysis may be overlooked or ignored. On the other hand, Alfred Marshall resisted mathematical expression of economic ideas because he sought to preserve the accessibility of economics from the vantage point of the common man.

Ideas that fuel economic analysis run parallel with the development of theory. The value of econometric and mathematical formalization in testing ideas is beneficial as long as limits to the final productiveness and efficacy of such techniques are understood. Econometrics has the potential to bog down economic studies in the tactical analysis of obscure ambiguities in the data collected instead of moving forward with the business of strategic problem solving. Some of the mathematical techniques developed recently have not necessarily been powerful or far reaching and so they are highly subject to revision and replacement. Ekelund and Hébert (1990) hold the view that as long the focus of economic study is on mathematics and empiricism, there is a potential in the debate to divide the discipline of economics into two separate camps; (1) economics as a branch of behaviorial science; (2) econometrics as a branch of applied mathematics.

The danger of creating a new discipline that focuses on the development of econometric techniques may be real, but that result would hardly be detrimental or without precedent. Varied schools of thought have been come and gone throughout the history of economic study. It is the very diversity of ideas which form the basis of a healthy and interesting dialogue between economists. Without a variety of ideas and great minds behind those ideas, the study of economics would be far less interesting and probably less relevant. With regard to the rich history of debate in economics, what is the harm in allowing the current fascination with econometric analysis to run its course?

Finally, over-formalization of theory in economics using mathematical or empirical technique could be in danger of rendering economics boring or irrelevant. Mathematical formalization should not be abandoned, but it should not be the driving factor in economic thought. The use of mathematics should be critically accepted with a judicious measure of skepticism. The runaway developments in econometrics may be slowing as some signs point to a slowdown in the production of articles about econometrics appearing in economic journals. It is important that those prominent in economics remember that absolute answers in economics are elusive and that mathematical techniques cannot find truth; they can only raise confidence, lower confidence or reduce ambiguity in postulated ideas. Indeed, the quest for scientific legitimacy has put the future of economics as a field of heterodox discourse in question.

Of course, one should not think that the use of mathematics is ruining the discipline of economics. Hardly. The availability of the many fine minds who are conducting research is remarkable and should have the power to tremendously improve our ability to interfere with dysfunctional aspects of the economy and leave alone those elements that are working fine. From a historical perspective, the message of the present fascination with econometrics is that it will shift and evolve into new solutions in ways that no one, not even the most astute historian of economics, can anticipate.

What is the end purpose of economic inquiry? Should not it be formulating practical answers to the questions of economic policy? This is not a new debate. In fact, the issue of mathematical formalization of economics has continually been debated since Adam Smith wrote The Wealth of Nations. It is in some sense a rather unusual turn of events though. Those who wanted to formalize economic theory have made great strides, but it was those thinkers such as Smith, Walras, Marshall, and Keynes who could think outside the box and break new ground who pushed the discipline forward. Is not it ironic that in a day and age when increased formalization of economics through computers is possible that some economists would cry, “Too much?”

Ekelund and Hébert (1990) raise several questions that are not easily answered in their conclusion to the chapter on The Development of Mathematical and Empirical Economics. Are graduate students of economics being trained improperly? Is the field of economics squandering its human capital and enamored with technique development but reticent to apply those same techniques in a cost-effective way? Do economists have low self-esteem about be treated as less than scientists? Can economics achieve the empirical validity of a true science? Can the development of econometric techniques move too far ahead of economic inquiry? Are professors and researchers of economics killing it as a discipline?

My view is that regardless of the weaknesses in some of their rhetoric, Ekelund and Hébert (1990) are essentially correct that too much formalization and verification in a field of study that must address real world, real time problems is harmful. The logical end point of the study of economics, by its very nature, is to understand economic behavior, and make practical suggestions, albeit imprecise ones, within a meaningful time frame. Otherwise, the modern economics has little practical value and could be in danger of becoming self-perpetuating, incestuous field of study.

Reference

Ekelund, R. B., Jr., & Hébert, R. F. (1990). A history of economic theory and method (3rd ed.). New York: McGraw Hill.