Every science derives hypotheses, generalisations, principles, laws and theories which explain the behaviour of phenomena it studies.
For deriving generalisations and theories it has to adopt a certain methodology.
One of the controversies in economics relates to the kind of methods to be adopted to discover generalisations and theories about the relationship between economic variables.
Economists have in fact adopted both deductive and inductive methods of reasoning. Recently econometrics, that is, application of statistical methods have gained much popularity among economists. These different methods will be explained below.Besides, the concept of equilibrium and role of equilibrium analysis in deriving economic theories will also be explained.
Further, the role of model building in economics to explain the behaviour of individuals and the economy will be brought out. The other aspect of the method or technique of economic analysis is whether it should be of the nature of statics, comparative statics or dynamics.
Nature of Scientific Theory:
A scientific theory sets up relationship between facts or, in other words, it explains cause and effect relationship between various variables. The variables with which economists are concerned are prices, quantities demanded and supplied, the money supply, national income, employment, wages, profits, etc. Every theory is based upon a set of assumptions, often called premises or postulates.
It is worth mentioning that some assumptions are taken merely to simplify the analysis, though they may not be entirely realistic. In economics, the assumptions may be behavioural, that is, relating to the behaviour of economic variables or they may be technological pertaining to the production technology and the availability of productive factors. From the assumptions or postulates some implications or conclusions are deduced through logical process of reasoning.
The process of logical deduction to discover relevant conclusions from a set of definitions and assumptions is carried out either in words or in the language of symbolic logic or it may be done with the aid of geometry or more formal mathematics. It is these conclusions drawn from the assumptions through deductive logic which are called hypotheses.
It is worth noting that a scientific hypothesis states the proposition about relationship between facts or variables in a form that is testable or falsifiable, that is, propositions which are capable of being refuted. If the predictions based on a hypothesis are refuted by the direct observation of actual facts or through the statistical methods of interpreting actual facts, a hypothesis stands rejected. If on testing the predictions based on a given hypothesis are proved correct, it stands established as a scientific theory.
For instance, the quantity demanded varies inversely with price is one of the important economic hypotheses established in economics. If a sales tax is imposed on a commodity and as a result the price of the commodity rises, the prediction will be that the quantity demanded will decline, other things remaining constant. This has not been falsified and in fact has been corroborated by the facts of the real world. So the law of demand stating that there is inverse relationship between price and quantity demanded is a scientific economic hypothesis.
Likewise, the generalisation regarding the direct relationship between price and quantity supplied, that is, the higher the price, the greater will be the quantity supplied, has also been found to be consistent with facts in several cases. Further, the Keynesian hypothesis that under conditions of less than full-employment of resources, levels of national income and employment are determined by the magnitude of aggregate effective demand is also a well established economic hypothesis regarding the developed capitalist economies.
The predictions based on the above Keynesian theory that the increase in aggregate demand through deficit budgeting by the Government under conditions of less than full-employment will lead to the rise in national income and employment has been generally found to be consistent with facts. Therefore, Keynesian theory of effective demand has been proved to be valid for the advanced capitalist economies by empirical evidence.
If the predictions based on a hypothesis are falsified by the facts of the real world, either some error would have been committed during the process of logical deduction or the assumptions made would have been too unrealistic, wrong or irrelevant to the subject of economic enquiry. Thus, in order to establish scientific economic hypotheses, error in logic and mistake of making unrealistic assumptions be avoided.
It is worth emphasising that every hypothesis or theory is based upon some simplifying assumptions which are not quite realistic, that is, it is an abstraction from reality.But good hypotheses and theories abstract from reality in a useful and significant way. Indeed, if we do not abstract from reality we would merely duplicate the real world in a camera like manner and will not gain any understanding of it. The crucial test of a hypothesis or theory is that whether predictions which follow from it are falsified or not by the empirical evidence, that is, by the facts in the real world.
If the predictions of the hypothesis or theory are found to be consistent with the facts, its assumptions would be justified even if they are unrealistic. Thus, a hypothesis or theory should not be criticised simply because assumptions it makes are unrealistic. The hypothesis is valid if its predictions are found to be consistent with the facts.
Derivation of Economic Theories and Nature of Economic Reasoning:
After having explained the scientific nature of economic hypothesis we are now in position to explain in detail how the generalisations in economics are derived and to clearly bring out the nature of economic reasoning. Economic generalisations describe the laws or statements of tendencies in various branches of economics such as production, consumption, exchange and distribution of income.
In the view of Robbins, economic generalisations and laws are statements of uniformities which describe human behaviour in the allocation of scarce resources between alternative ends. The generalisations of economics like the laws of other sciences, state relationship between variables and describe those economic hypotheses which have been found consistent with facts or, in other words, have been found to be true by empirical evidence. But a distinction may be drawn between a generalisation (law) and a theory.
A generalisation or law just describes the relationship between variables; it does not provide any explanation of the described relation. On the other hand, a theory provides an explanation of the stated relation between the variables, that is, it brings out the logical basis of the generalisation. An economic theory or a model derives a generalisation through process of logical reasoning and explains the conditions under which the stated generalisation will hold true.
Generalisations in economics have been derived in two method:
(1) Deductive Method,
(2) Inductive Method.
1. Deductive Method:
We shall first explain the deductive method of deriving economic generalisations. The deductive method is also called abstract, analytical and a priori method and represents an abstract approach to the derivation of economic generalisations and theories.
The principal steps in the process of deriving economic generalisations through deductive logic are:
(a) Perception of the problem to be enquired into;
(b) Defining precisely the technical terms and making appropriate assumptions, often called postulates or premises;
(c) Deducing hypotheses, that is, deriving conclusions from the premises through the process of logical reasoning; and
(d) Testing of hypothesis deduced.
(a) Perception of the Problem:
In any scientific enquiry, the analyst or theorist must have a clear idea of the problem to be enquired into. He must know the significant variables regarding whose behaviour and interrelationship he wants to derive generalisations. The perception of the, problem is by no means an easy task.
(b) Definition of Technical Terms and Making of Assumptions:
The next step in the process of deriving generalisations is to define precisely and unambiguously the various technical terms to be used in the analysis as well as to state clearly the assumptions or postulates one makes to derive generalisations. As mentioned above, assumptions may be behavioural pertaining to the behaviour of the economic variables or they may be technological relating to the state of technology and the factor endowments.
The crucial assumptions are made on the basis of observations or introspection. A crucial assumption that has been made in economics is that consumers try to maximise their satisfaction and producers try to maximise their profits. Likewise, it is assumed that investors try to minimise their risk and maximize the expected rate of return. Some of the assumptions are made merely to simplify the analysis and may not be quite realistic.
The actual economic world is quite complex and full of details in which numerous factors play a part and act and interact on each other. The introduction of simplifying assumptions is quite necessary in order to bring out the importance of really significant factors having a bearing on the problem under investigation.
According to Prof. Boulding, economic theory represents just a ‘map’ of real world phenomenon and not a perfect picture of it. To quote him, “Just as we do not expect a map to show every tree, every blade of grass in a landscape, so we should not expect economic analysis to take into account every detail and quirk of real economic behaviour.
It, therefore, follows that each and every assumption made by a theory may not be realistic. The crucial factor in building up a valid theory is whether its predictions are corroborated by the facts in the world. A correct scientific theory or generalisation must be expressed in the form of a hypothesis that is conceivably refutable. As mentioned above, Professor Friedman in his now well-known article, “The Methodology of Positive Economics” has expressed the view that undue importance should not be given to the ‘realism’ of assumptions. What matters most from the viewpoint of scientific theory, according to him, is whether it enables us to predict accurately.
(c) Deducing Hypotheses through Logical Deduction:
The next step in deriving a generalisation through deductive logic is deducing hypothesis from the assumptions or premises taken. An hypothesis describes the relationship between factors affecting a phenomenon; it establishes cause and effect relationship between the variables having a bearing on the phenomenon. Then through logical process, hypothesis is deduced from the assumptions made.
This logical reasoning may be carried out verbally or it may be conducted in symbolic terms using the language of what is known as symbolic logic. The geometric or graphic technique is also usually employed to deduce the hypothesis about the relationship between factors. Besides, the process of logical deduction may be done with the help of more formal mathematics.
Nowadays in almost all branches of modern economics, the mathematics as a tool of analysis for deriving economic theories and generalisations is being increasingly used. The use of mathematics in economic analysis proves extremely useful where geometrical methods make the analysis more complicated to comprehend. Besides, the use of mathematical method makes the derivation of economic hypotheses more rigorous and exact.
It is worthwhile to note that in deriving analytically sound hypotheses, one should guard against committing logical fallacy in the process of logical deduction. For instance, it is inappropriate to conclude that A must be the cause of B, if A happens to precede B. Further, it is logically fallacious to argue that since there exists a high degree of correlation between the two factors, say between the supply of money and the general price level, the former must be the cause of the latter, unless the causation must be logically developed.
(d) Testing or Verification of Hypotheses:
Hypotheses obtained above have to be verified before they are established as generalisations or principles of economics. For the verification of hypotheses, economists cannot make controlled experiments, because they have to discern uniformities in behaviour patterns of man.
As we cannot make experiments with man under controlled conditions, such as in laboratories as physical scientists make experiments with inanimate objects of nature and biologists make these with animals and plants. Therefore, economists have to rely on uncontrolled experiences and observations. This information regarding uncontrolled experiments about the behaviour patterns concerning variables about man and the economy are quite amply available.
The reliance by economists on uncontrolled experiences, however, does increase the number of observations required to verify the hypotheses or to establish the generalisations. Besides, the need to rely on uncontrolled experiences complicates the analysis and requires that facts must be carefully interpreted to discern successfully the significant relationship between relevant economic variables. Prof. Baur rightly remarks.
“The need to rely on uncontrolled experiences does, however, increase the number of observations required, and also complicates their successful analysis and interpretation, before we can discern successfully the significant uniformities and ascertain their limits.” He points out that in spite of the complexities and difficulties involved in verifying economic hypotheses through successful analysis and proper interpretation of uncontrolled experiences and observations, several useful and significant generalisations have been established in economics.
In the field of microeconomics, the well-established generalisations relate to the inverse relationship between price and quantity demanded, the direct relation between price and quantity supplied, the tendency of the price of the product to be equal to the marginal cost under conditions of perfect competition, and the tendency for the wages to be equal to the value of marginal product under conditions of perfect competition and several others.
In the field of macroeconomics, established generalisations relate to the determination of the level of national income by aggregate demand and aggregate supply in a capitalist economy, the multiple increase in income and employment as a result of a given initial increase in investment depending upon the size of marginal propensity to consume, the dependence of the amount of investment on the marginal efficiency of capital and the rate of interest and several others.
“It is worth noting that the absence of controlled or contrived experiments in economics affects the forms of various generalisations in different degrees.” This means that the generalisations in economics are not as exact as those of physical sciences and they are therefore not universally applicable under all circumstances. Because of the absence of contrived experiments economic generalisations lack in firmness. Economic generalisations are therefore not easily accepted by all.
Even generalisations that are refuted by empirical evidence are not abandoned for good by all. Prof. Baur rightly points out, “the absence of the vivid and dramatic evidence provided by the contrived experiments adds greatly to the difficulty of securing acceptance for generalisations which are amply justified by the analysis of the available evidence.” Likewise, absence of controlled experiments, according to Friedman, “renders the weeding out of unsuccessful hypotheses slow and difficult. They are seldom downed for good and are always cropping up again.”
In regard to framing and testing of economic generalisations, two related distinctions must be borne in mind. First, functional relationship between economic variables and a historically sequence of events must be distinguished. For instance, the law of demand stating inverse relationship between price and quantity demanded does not become invalid in view of the fact that both prices and quantities sold of many commodities increase during a boom period.
This is because certain other forces such as a rise in aggregate investment demand operates which causes increase in both the price and quantity sold during a boom period. Second, predictions of a generalisation to show its validity must be carefully differentiated from the forecasting of future events; actual events may not exactly come about as predicted by a generalisation and yet that generalisation may be correct.
This is because, as mentioned above, the actual course of events is governed by several other factors assumed by a generalisation which remains constant under the qualification “other things remaining the same”. Thus, “even if the prediction that producers of a particular crop respond to a higher price by producing more is correct, this prediction does not enable us to forecast accurately next year’s output (still less the harvest in the more distant future), which in the event will be affected by many factors besides changes in price.”
It follows from above that in the absence of controlled experiments, for the verification of their generalisations economists have to rely on the direct observations of the events in the real world. By direct observations we mean “gathering of information personally or reliance on comparatively unprocessed material such as files of business firms and government departments, locally published reports, proceedings of representative assemblies, newspapers, advertisements, market reports, auction notices and the like.”
In order to prove the validity of hypotheses and therefore to establish generalisations, importance of direct observations cannot be underrated. Thus Prof. Baur assets, “The depth and significance of economic generalisations depend on the quality of the underlying observations and analysis.”
Testing of Economic Hypotheses through Econometrics:
In recent years a very useful method to test economic hypothesis has been developed. This is the statistical method or what is now popularly called econometric method. The statistical or econometric method to verify and establish the theoretical generalisations occupies an important place because there is limited applicability of controlled experimentation in economics.
The various statistical methods such as regression analysis have been developed to empirically test the economic hypotheses on the basis of collected economic data. The merit of econometrics is that the degree of functional relationship between relevant economic variables in precise quantitative terms is obtained by it and also the level of significance of the results can also be estimated.
Recently, econometric method has been used to establish the precise relationships between money supply and the price level, quantity of money and the national income, consumption and income, capital accumulation and rate of economic growth and so forth.
It may, however, be pointed out that statistical analysis or econometrics alone cannot be used to derive and establish economic principles and theories. Economic hypotheses or theories must be developed logically before we can meaningfully use statistical analysis to test and verify them. Indeed, theory or hypothesis is needed before the selection of the relevant facts and data regarding relevant variables which can be subjected to empirical testing through the methods of econometrics.
Prof. Myrdal is quite right, when he says, “Theory, therefore, must always be a priori to the empirical observation of facts. Facts come to mean something only as ascertained and organised in the frame of a theory. Indeed, facts as part of scientific knowledge have no existence outside such a frame. Questions must be asked before answers can be obtained and, in order to make sense, the questions must be part of a logical coordinated attempt to understand social reality as a whole. A non-theoretical approach is, in strict logic, unthinkable.”
Merits and Demerits of Deductive Method:
The deductive approach to establish economic generalisations was extensively used by Classical and Neo-Classical economists such as Ricardo, Malthus, Senior, J.S. Mill, Marx, Marshall and Pigou. It still remains popular with modern economists as it has several merits. First, useful mathematical techniques can be employed to derive generalisations of economics.
With the aid of rigorous mathematical logic, economic theories can be developed through the process of deduction which can successfully explain economic phenomena. Secondly, through deductive logic useful economic theorems can be derived without the tenuous and detailed collection and analysis of data which are required under the alternative inductive method.
Thus, as compared to inductive method, method of deduction is less time-consuming and less expensive. Thirdly, in view of the limited scope for controlled experimentation in economics, the method of deduction is an extremely useful method of deriving generalisations. This is because multiplicity of forces acts simultaneously on an economic phenomenon and it is not possible to eliminate some of these by means of a controlled experiment.
This indicates the crucial importance of deductive logic for building up economic principles or generalisations. Fourthly, the use of sophisticated mathematical methods in the deductive approach enables the economists to introduce accuracy and exactness in economic principles and theories.
In spite of the above merits, shortcomings of the deductive approach should not be overlooked. The use of deductive method in deriving economic generalisations requires the use of a high-level competence in logic and theoretical abstraction. A good deal of care and objectivity is needed to avoid bad logic or faulty economic reasoning. Prof. Blaug rightly opines, “It is perfectly true that economists have often deceived themselves and their readers by engaging in what Leontief once called “implicit theorising” presenting tautologies in the guise of substantive contributions to economic knowledge.”
Besides, most economists have preconceived notions or biases on several economic issues. If sound and valid economic generalisations are to be established, economists must dissociate themselves from normative preconceptions and biases in their logical process of deducing valid economic generalisations.
Further, a great demerit of deductive approach is that with it highly sophisticated theoretical models based on highly unrealistic assumptions may be developed which do not have any operational significance. Indeed, such highly irrelevant analytical models with little empirical content and incapable of being used for policy formulation have in fact been developed by economists. Such models are no more than mere “intellectual toys”. If economics is to serve as an instrument of social betterment, building of such theoretical models having no operational use should be avoided.
Lastly, in the derivation of economic hypotheses and conclusions through deductive logic, assumptions play a crucial role. If the assumptions made are such that when on removing them, economic hypothesis based on them is refuted, then making of these assumptions is not valid. Thus, one who uses deductive approach should always keep in mind to what extent the validity of generalisations derived depends on the assumptions made.
For instance, the Keynesian macro-analysis is based upon the assumption of a depression-ridden capitalist economy with a lot of excess productive capacity. Therefore, a positive harm has been done in applying the Marxian theories in the context of developing countries such as ours where the assumptions made by Marx do not hold good. Hence, mere “deductive arm-chair analysis” should be avoided, if the scientific character of economics is to be maintained.
2. The Inductive Method:
The indicative method which is also called empirical method derives economic generalisations on the basis of experience and observations. In this method detailed data are collected with regard to a certain economic phenomenon and effort is then made to arrive at certain generalisations which follow from the observations collected. But, it is worth mentioning that the number of observations has to be large if it can yield a valid economic generalisation. One should not generalise on the basis of a very few observations.
There are three ways which can be used for deriving economic principles and theories. They are:
(c) Statistical or econometric method.
The experimentation, that is, the use of controlled experiments is of limited applicability in economics. First, unlike natural sciences which are concerned with analysing the behaviour of either inanimate objects or obedient animals such as rats and rabbits under the influence of chloroform, economics deals with the behaviour of man who is quite fickle, sensitive, wayward and unmanageable.
Besides, man cannot tolerate the idea of being experimented upon, either individually or collectively. Further, an economic phenomenon is the result of multiplicity of factors and causes acting and inter-acting upon each other. Therefore, economic phenomenon does not repeat itself in the same uniform pattern. Numerous factors acting on an economic phenomenon ‘disturb’ it and make its exact repetition unlikely.
Thus, as compared with the natural phenomena, economic phenomena are of less uniform pattern, less repetitive and more variable. Furthermore, economists study the economic phenomena in which pressure groups such as employers’ associations, trade unions, farming lobbies, political parties with their different ideologies play a crucial part and their activities render it difficult to make controlled experiments in the economic world.
However, in spite of these difficulties, experimental method can be used in some fields. For instance, experiments have been conducted to find out which law of production is valid, that is, whether law of diminishing returns, law of constant returns or law of increasing returns operates in the real world. Besides, public undertakings or big industrial firms often try to assess the effect of the changes in the prices of their products on the demand for it and thus find out the demand elasticity of their products.
As has been explained above, observations of facts through collection of detailed data and the use of statistical methods to arrive at economic generalisations describing relationship between facts are being increasingly made. Some of the recent researches in the field of macro-economics, such as the nature of consumption function describing the relation between income and consumption, the principle of acceleration describing the factors which determine investment in the economy have been obtained through the use of mainly inductive method.
However, it needs to be emphasized again that the use of induction or empirical method is not of much value if it is not supported by the economic hypothesis or theory developed by deductive logic. The inductive or statistical method can at best be used to empirically test the theory or hypothesis as to whether it is consistent with or refuted by facts.
The inductive method has another limitation in that there is a great risk of conclusions being drawn from insufficient data. To obtain generalisations through empirical method, one should take care that sufficient number of observations or data has been taken into account. Besides, the collection of data itself is also not an easy task.
And a researcher who wants to employ the inductive method to arrive at generalisations must have good knowledge of statistical methods, that is, he must know the art of collecting, processing and interpreting data. It is obvious that as compared with the deductive method, the inductive method is time-consuming and expensive.
Conclusion: Integration of Two Methods:
Now, the controversy which existed among the earlier economists as to whether deductive or inductive approach is more appropriate in developing economic theories and principles has been resolved. The modern viewpoint in this regard is that both are needed for the proper development of scientific economic theories. Indeed, the two are complementary rather than competitive.
The modern economists first derive economic hypotheses through the process of logical deduction and then empirically test them through statistical or econometric methods. Marshall rightly pointed out, “induction and deduction are both needed for scientific thought as the right and left foot are both needed for walking.”
Empirical studies made through statistical or inductive method without a theoretical hypothesis to serve as a guide for the selection of data are quite useless. The derivation of economic generalisations through the approach of deductive logic without empirically testing them through inductive method is also not quite proper. Empirical studies made in inductive approach also bring to light significant economic facts or phenomena which require analytical explanation through deductive logic.
For instance Farm Management Studies in India in the mid fifties led to the discovery of a fact that output per acre on the small-sized farms is higher than that on large farms. This led to the various theoretical explanations of the phenomenon observed in the empirical studies.
On the other hand, a theory or hypothesis is first developed through deductive logic from some assumptions and then predictions based on the hypothesis are tested through inductive or statistical method. If the predictions are found to be consistent with facts, the hypothesis or theory stands proved and if the predictions of the theory are found to be inconsistent with facts, it stands rejected.
Role of Assumptions in Economic Theory: Friedman’s View:
As has been pointed out above, every law and generalisation of economics is based upon some assumptions. Now, the question is whether for the formulation of proper economic laws these assumptions should be realistic or not. One view is that laws of economics if they are to be realistic must be based upon assumptions which are realistic too.
Thus, according to this view, making unrealistic assumptions and establishing laws on their basis will make these laws invalid. However, a contrary view has been put forward by Prof. Milton Friedman of Chicago University in his now well-known article, “The Methodology of Positive Economics.” In this context Prof. Friedman draws a distinction between positive economics and normative economics. According to Prof. Friedman, positive economics explains “a system of generalisations that can be used to make correct predictions about the consequences of any change in circumstances. Because the predictions of this positive economics have to be tested with the empirical evidence it is as much a science as any other physical science even though the assumptions made may be unrealistic.
The crucial point is whether the predictions based on economic generalisations of positive economics are confirmed by the facts and empirical evidence. According to Friedman, assumptions cannot be realistic; since they are made merely to simplify the analysis.
However, it may be pointed out that while drawing conclusions from economic theories and laws regarding economic policies it must be known whether the assumptions made do not make the policy conclusions invalid if these assumptions are removed. Dr. K. N. Raj has rightly said, “Some of the differences between economists of policy questions can be traced to the assumptions they choose to make when faced with the problems of this kind.” He further adds, “it is, however, essential in the interest of clarity and intellectual honesty that economists state clearly the assumptions on which one set of policies and programmes is advanced in preference to another and the reasons for making these assumptions”