NARENDAR PANI

T

Role of Intuition

Recognising this role of intuition is undoubtedly a major step forward. It takes the debate beyond an objective method to the actual practice of economics where these objective tools are only a part of the entire process. But an explicit recognition that intuition has a role to play only throws up other questions. Is the role of intuition so insignificant in the entire process that it can be ignored? And, if it is not, is intuition unbiased? This note argues that the answers to both these questions are in the negative.

Most dictionaries attribute two characteristics to “intuition”, immediacy and the lack of reasoning. A typical example is the Concise Oxford Dictionary definition of intuition as “immediate apprehension by the mind without reasoning” (COD 1990: 623). When economists use the term intuition, however, it is not always clear that both these characteristics are being implied. Immediacy, in particular, is not an essential part of the judgment a development economist would make when deciding whether a result that holds for one population should be extended to another. The development economist’s decision may be carried out without reasoning or the articulation of reasoning, but here again, we must guard against giving the impression that no thought goes into the process. What has been referred to as intuition may then perhaps be better described as a subjective judgment.

The extent to which this subjective judgment is used in practice is much more substantial than is generally accepted. The very choice of the problem the development economist chooses to study is not devoid of all subjective judgment.

It can be argued that the problems can themselves be thrown up by data. The first stage in Mookherjee’s research process in fact includes, “exploratory data analysis aimed at helping identify empirical regularities that need to be explained by a suitable theory” [Mookherjee 2005]. But the decision of where and on what issues to carry out such exploratory data analysis remains a subjective judgment. Among other things, a researcher faces the choice of following the issues that at that point of time have the greatest attention of her peer group, or simply going out in a completely new direction on the basis of nothing more than a personal whim. While the first choice may have better chances of providing a deeper understanding of existing areas of interest, not to mention a greater recognition by the peer group, the latter may lead to fundamentally new research. Which route the researcher chooses is essentially a subjective judgment.

Scope of Subjective Judgments

The scope for such subjective judgments can be even greater when researchers try to put together a series of microscopic results in order to capture the larger picture. The researcher has to decide just how much importance each result will get and how exactly it will fit into the larger picture. This problem is not confined to development economics. When the Asian crisis unfolded in 1997, major economists of otherwise different persuasions agreed that they had not anticipated it. Even critics of the earlier south-east Asian performance, like Paul Krugman, did not quite expect the crisis to take the form it did. While they expected a conventional currency crisis, what they got was a crisis that did not follow the patterns laid out by standard currency crisis models. There was no fiscal imbalance on the eve of the crisis, the inflation rates were low, there was no substantial unemployment and the economies seemed to have already completed a boom-bust cycle. And yet, it did not take very long for economists to come up with precise explanations for the various dimensions of the Asian crisis. What is more, many of the individual concepts used to explain the crisis were not new. Competitive devaluation, crony capitalism, moral hazard, panic in the markets and many of the other concepts were not unknown before June 1997. What was new was the manner in which individual economic processes worked together. No major economist had guessed that these processes would interact with each other in exactly the way that would cause the Asian crisis. And there is no reason why the Asian crisis should be treated as an exception. As Krugman has pointed out, “As is all too often the case, we find ourselves playing theoretical catchup – trying after the fact, to develop a framework for

by the experience and quality of these economists. The bias in subjective judgments can thus be of very different kinds, and have both desirable and undesirable elements.

Appraisal and Characterisation

In developing a response to subjective judgments with a variety of not unbiased elements, it may be useful to tap Ernest Nagel’s distinction between characterising and appraising value judgments. Characterising value judgments are evaluations of evidence, “which conclude that a given characteristic is in some degree present (or absent) in a given instance” [Nagel 1984: 492]. Such characterising subjective judgments are more widespread in the practice of development economics than is often explicitly recognised. Statistical tests are expected to tell us whether a particular characteristic exists in a population or not. But the researcher does need to make a judgment about what she considers sufficient evidence to believe a particular characteristic exists. Should she accept a 5 per cent degree of significance or insist on a 1 per cent degree of significance? The pursuit of rigour may recommend an insistence on a 1 per cent degree of significance. But this implies that it is better to reject a pattern that might be true rather than accept one that is likely to be false. And this itself is a subjective judgment.

Appraising value judgments, on the other hand, are evaluations “which conclude that some envisaged or actual state of affairs is worthy of approval or disapproval” (ibid, p 493). The choice of desirable outcomes for development policy would itself involve such judgments. Development economists may be unanimous on the need to remove poverty, but the very idea that poverty is undesirable is itself an appraising subjective judgment.

Since both appraising and characterising subjective judgments exist, it is important that they are made explicit so that they can be subject to a closer scrutiny. And the measures needed to evaluate the quality of appraising judgments could be quite different from those required to evaluate the quality of characterising judgments. As far as appraising judgments go, once what is considered desirable is made explicit, it should not be difficult to see the moral tradition from which these judgments are being made. It will then be quite evident whether or not the moral compass of the development economist is in tune with those that society at large demands. What is important here is not so much the scope for subjective judgments as whether it is consistent with what society considers desirable.

The task of evaluating characterising subjective judgments is rather more difficult, as these subjective judgments are being made in pursuit of an objective truth. The subjective judgment only goes so far as to decide whether the evidence is sufficient or not to believe a particular characteristic exists. The answer to this question would be best if the role of subjective judgments is minimised. But, to the extent that characterising subjective judgments remain, it becomes necessary to evaluate the development economist’s belief that a certain amount of evidence is sufficient to hold that a certain characteristic exists. It is here that a return to some of Keynes’ work on probability might help.

Keynes had used probability to understand issues on which there could only be partial belief. It is now recognised that it “is the essence of Keynes’s theory that is alive today; the detailed way in which he worked it out has not survived” [Braithwaite 1973, p xvi]. The essence of the treatise has largely survived by shifting the focus from the probability relationship to the consequences of that probability, that is, the degree of belief. As Braithwaite puts it:

Many of those today who think about the logic of partial belief … would start with the degree of belief and consider what conditions this must satisfy in order to be regarded as one which a rational man would have under given circumstances. To start this way requires a notion of degrees of belief which is independent of considerations of rationality, and this was provided by F P Ramsey in a paper of 1926 written deliberately as a constructive criticism of Keynes’s view. Ramsey proposed to measure the degree of belief in a proposition which a particular man has at a particular time by the rate at which he would be prepared to bet upon p being true, which is to say that a belief of degree q (with 0 ≤ q ≤ 1) if the man is prepared to pay a proportion q of one unit of value (but no more) for the right to receive one unit of value if p is true but nothing if p is false. Degrees of belief when measured in this way will be called (following Rudolf Carnap) betting quotients: A betting quotient measures a man’s actual partial belief at a point of time, and in no way depends on whether or not the man has a good reason for holding a partial belief of that degree [Braithwaite 1975, pp 240-41].

This measure of belief would be particularly useful in evaluating characterising subjective judgments in questions involving prediction. The measure could then be used when comparing the prediction with the actual result. There is, of course, no guarantee that an economist who has got it right in the past will do so in the future as well, or vice versa, but there is likely to be a greater degree of confidence in the characterising subjective judgments of an economist who has a successful record. It could be argued that a similar process is already evident in the mainstream economic policy-making. Alan Greenspan’s record in the Federal Reserve did improve the confidence in his ability to judge what exactly was going right, or wrong, with the American economy. Betting quotients would only extend what is currently confined to an Alan Greenspan to the larger body of economists in a manner that is more rigorous than relying on popular imagery.

This would still leave a number of unanswered questions such as, how do we evaluate the quality of a characterising subjective judgment when the statement being made does not have a predictive dimension to it? But what should be clear is that subjective judgments in the practice of economics call for a more detailed response than merely acknowledging a role for intuition.

EPW

Email: narendar.pani@gmail.com

References

Bardhan, Pranab (2005): ‘Theory or Empirics in

Development Economics’, EPW, October 1. Basu, Kaushik (2005): ‘New Empirical Develop

ment Economics: Remarks on Its Philosophical

Foundations’, EPW, October 1. Braithwaite, R B (1973): ‘A Treatise on Probability’,

The Collected Writings of John Maynard Keynes, Vol VIII, Macmillan Press, London.

– (1975): ‘Keynes as a Philosopher’ in Milo Keynes (ed), Essays on John Maynard Keynes, Cambridge University Press, Cambridge, pp 237-46.

Krugman, Paul (1998): ‘What Happened to Asia’,

paper presented to a conference in Japan in

January 1998, downloaded from the internet. Mookherjee, Dilip (2005): ‘Is There Too Little

Theory in Development Economics Today?’,

EPW, October 1.

Nagel, Ernest (1984): The Structure of Science: Problems in the Logic of Scientific Explanation,

Macmillan, Delhi, India.