

Table: Terminology in positive economics 

The essence of
scientific work The covering law model suggests that the production of scientific
theories involves two types of work: induction and deduction. 

Lead to. Implies
Starting from a set of observations one generates (explanans) a set of
premises. More to the point, induction is the process by which the
researcher is able to generate the explanans by
observation of facts. For instance, much of the work economists devote to
case studies, data collection, and statistical analysis is part of the
inductive process. Economics allows for two kinds of induction empirical
induction and hypothetical induction: 


Empirical induction is the most demanding and scientific of the two kinds of induction. It requires the researcher or scientist to produce as accurate an explanans as possible. Explanans consists of 1) sentences describing fundamental behavior or action, and 2) sentences describing the conditions under which this type of behavior applies. As a result, a ‘well done’ empirical induction implies making an accurate description of the fundamental behavior, including an exact description of the conditions limiting the domain of this behavior. The terms; accurate and exact, should be understood relative to the facts that are studied. In other words, empirical induction does not allow the explanans or part of it to be without empirical content. 

Hypothetical induction is much less restrictive than empirical induction. It is used as a best but inferior alternative in situations where the studied phenomena is too complex to be formulated in terms of mathematics, and at the same time to be a realistic representation of reality. In these situations, the researcher is forced to tradeoff less realism with more simplicity in order to preserve the mathematical approach. This is done by hypothetical induction. This kind of induction allows large parts of explanans to be without empirical content, although some of it must be empirical in order to maintain a scientific status. For instance, mathematical models in economics typically contain a few realistic assumptions and a bunch of simplifying assumptions. Individually, many of these simplifying assumptions do not appear unrealistic, but taken together it is often unrealistic to imagine any real world situations where all the conditions are satisfied. From a scientific point of view this is clearly nonsatisfactory since it prevents us from making statements of direct, empirical relevance. Nevertheless, such models are worthy because of their huge indirect, empirical relevance (A link about this is under construction). Hypothetical induction even allows the researcher to include propositions that are counterfactual. For instance, almost all mathematical models in economics assume that the economic players are perfectly rational. This is wildly unrealistic from a physiological point of view, but its inclusion is necessary for reasons of simplicity. 

Deduction means, lead away from, in the sense of starting from a set of
premises, one derives a conclusion.
In other words deduction is the
process by which the researcher uses his knowledge about explanans
to generate explanandum. Economists typically
apply mathematical or intuitive procedures to generate the lower order
behavior (explanandum) from the fundamental (or higher order) behavior
described in explanans. Two kinds of deduction can be found in economics:
logical deduction and intuitive deduction. 

The most scientific and classic type of deduction is logical deduction. This kind of deduction requires the researcher to express explanans in terms of a mathematical model, and then to derive explanandum by algebraic manipulations of this model. The advantage of applying mathematics is to minimize the chance of making errors and to ‘save on human rationality’. The errors are minimized because the mathematical language is far more exact and correct than the verbal or spoken language. Furthermore, the mathematics is able to economize on our limited rationality, because it enables us to make successive manipulations of formulas until a concluding formula emerge. Concluding formulas are often nonobvious and are therefore unlikely to be revealed by use of pure intuition. A metaphor may help to explain what is meant: A mathematician is like a great warrior, a Samurai, who can win a war (derive a formula) over 20 peasant warriors if he is allowed to fight them one by one (if the mathematician is allowed to make 20 sequential algebraic manipulations). However, the Samurai would certainly be killed if he should fight 20 men at the same time (the mathematician would fail to derive a formula if he was denied pen and paper on which to make successive manipulations). 

Sometimes it is hopeless to state the explanans in terms of mathematics and as a result logical (or mathematical) deduction is impossible. In this case, the best alternative is to use intuitive deduction. This kind of deduction implies, that the researcher uses his common sense to figure out what conclusions seem to be the most likely given the available formulation of explanans. Relative to algebraic manipulations, intuition or common sense is a very subjective and erroneous kind of deduction, and from a scientific point of view this is a huge problem. However, it is still far better than having no theories at all to explain a phenomenon. 

The basis of a
scientific theory: From the point of view of positive science the basis of a scientific
theory is 1) the facts / reality it is supposed to describe, 2) the explanans or
the premises of the theory, and 3) the explanandum or the
conclusions of the theory: 

Scientific theory is
always inspired and based on observations of reality, because it is reality
that any scientific theory tries to describe or disclose. Both the explanans and the explanandum
of a theory should be related to facts or reality. However, in economics it
is especially important that the explanandum (the predictions of the theory)
are confirmed by facts / reality. This is one of the key messages by
Friedman's [1953] influential paper on the methodology of positive economics.
Indeed, this paper argues that as long as explanandum is confirmed by facts a
simple but unrealistic formulation of explanans may be preferred to a
realistic but complex formulation of explanans. The corporate governance
encyclopedia refers to such theory as predictional
theory because it focuses on the ability of theories to predict. Machlup
[1955] takes another point of view and argues that both the explanans and the
explanandum should resemble reality even if it comes at the price of
scarifying simplicity. Such theory is referred to as explanatory theory in this encyclopedia
because it stresses the importance of credible explanations (explanans) to
observable phenomenon. 

The explanans is the premises of a theory, or “the class of those sentences which are adduced to account for the phenomenon”, Hempel and Oppenheim [1948, p. 321]. To be more specific, explanans is composed of two kinds of sentences: 1) sentences or formulas describing some kind of fundamental behavior or action, and 2) sentences describing the conditions under which this kind of behavior applies. In economics an example of fundamental behavior could be utility maximization or risk aversion, and examples of conditions could be asymmetric information or monopoly. Explanans is also sometimes described as 'that with which we explain'. 

The explanandum is the conclusion of a theory, or “the sentence describing the phenomenon to be explained (not that phenomenon itself)”, Hempel and Oppenheim [1948, p. 321]. We may also say that explanandum is an expression of behavior, but of a lower order than the behavior in explanans Braithwaite [1953, page 12] states it elegantly: “The propositions in a deductive system may be considered as being arranged in an order of levels, the hypotheses at the highest levels being those which occur only as premises in the system, those at the lowest level being those which occur only as conclusions in the system, and those at the intermediate levels being those which occur as conclusions from deductions of higherlevel hypotheses and which serve as premises for deductions of lowerlevel hypotheses”. Click here to see an exhibition on the many lower level hypotheses in corporate governance. Explanandum is also sometimes described as 'that which should be explained'. 

The three
archetypes of scientific theories This encyclopedia distinguish between three archetypes of scientific
theories: 1) Explanatory theory as suggested by Machlup
[1955]. 2) Predictional theory as suggested by Friedman
[1953]. 3) Exemplary theory that combines the most
scientific attributes of theory production. 

Explanatory theory that uses empirical induction and intuitive deduction. This is theory that
tries to establish as realistic an explanans as possible, but because of this
realism the explanans is too complex to be formulated mathematically and the
explanandum is therefore derived by intuitive means. Coase, Hayek, and Simon
are some of the less numerous Nobel Laureates that are honored for theories
of the explanatory kind. The name of explanatory
theory is due to the theory’s prime interest in the explanans of a theory,
and the function of explanans is clearly to explain explanandum. 

Predictional
theory uses hypothetical
induction and logical deduction. This is
theory that includes a number of highly unrealistic but simplifying
assumptions. As a result of this simplification the explanandum is capable of
being deduced mathematically. Mirrless, Modigliani, and Sharpe are just a few
of the many Nobel Laureates in economics that are awarded for this kind of
theories. Predictional theory is called so, because
its main interest is the explanandum of a theory, and this is predictional in
the sense that explanandum is an induced type of behavior, i.e. in principle
it is not known in advance. 

Exemplary theory refers to theory that uses empirical induction and logical deduction. This is theory that not only is advantaged by having a highly realistic explanans but also benefits from being able to deduce its explanandum by mathematical means. Unfortunately, this kind of theory is practically impossible in the social sciences (a link to a thorough discussions of the reasons will be added soon). However, in the natural sciences exemplary theory is the most common way of doing science. 

 Copyright 19972018, H. Mathiesen. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Legal notice. 
.