The Satire of Econometrics

Modelling Money Revolution in Darjeeling

This article explores various principles of rationality and decision-making, analysing how these concepts apply to modelling economic behaviour and revolution. It discusses the maximum principle, preference relations, and the implications of transitivity in decision-making processes, with a focus on how these theories can be interpreted in the context of Darjeeling's economic landscape

All principles of decision-making under Political-Economy uncertainty justify revolution, which renders the system rational. Maximisation is fundamental to ensuring a sensible person's reasonable security level. Consider someone aiming to be a neta; what is required for ‘rationality’? Hamburger (1979, p. 17) provides a clear explanation through a coin-picking game that employs the concept of dominance to describe the decision-making scenario. The essence of these examples is to illustrate ‘vulnerability’.

Now, let us examine the decision principle (see, for example, Giere 1979: 337); one might adopt a principle focusing on the worst possible outcome from any option under all conceivable circumstances and choose the option that provides the greatest benefit under the worst scenario, which is commonly known as the ‘maximum principle’ or ‘play-it-safe rule’. This implies that the maximum probability of vulnerability results in the minimum payoff.

This essay explores various avenues through which rationality contributes to the growth of rational wealth maximisation, assuming S, B1, B2, A, E, M, etc., are the variables summing to GL.

Some of these principles may seem confusing, intentionally so, but they provide a robust foundation for rational behaviour under specified conditions. These principles are all part of the utility maximisation rules (see Harsanyi 1977: 22–47). To clarify, utility maximisation under certainty, risk, and uncertainty can all be regarded as rational choice principles. Rationality here is understood in a ‘thin’ sense; that is, behaviour is rational if it aligns with the actor’s preferences under scrutiny. For instance, if A is making a rational choice, A will choose M over B whenever A prefers M to B (B1, B2). If A is indifferent between M and B (B1, B2), A is equally likely to choose either.

To illustrate further, consider A faced with a choice between M and B (B1, B2) and examine what conditions A's preferences must meet for us to judge A's behaviour as rational. A must establish a preference relation between M and B (B1, B2). In other words, A must be able to determine whether M is preferred to B (B1, B2), denoted by M > B (B1, B2), or B (B1, B2) is preferred to M, denoted by B (B1, B2) > M, or A is indifferent between M and B (B1, B2), denoted by M ~ B (B1, B2). Technically, we must assume that A's preference relation over options M and B (B1, B2) is complete (or connected). If none of these possibilities hold, we cannot assert that A's choice aligns with A’s preferences.

Another requirement is the asymmetry of strict preferences. If M is preferred to B (B1, B2), then B (B1, B2) must not be preferred to M. Conversely, indifference should satisfy the symmetry condition: if A is indifferent between M and B (B1, B2), then A should be equally likely to choose either option with no preference.

Suppose the conditions are met in the choice scenario involving M and B (B1, B2). In that case, we can assign actual values to M and B (B1, B2) so that, when choosing between these two, the GL behaves in accordance with these variables to maximise the numerical value. For example, if A prefers M to B (B1, B2), we might assign M a value of 100 and B a value of 10. If picking M aligns with A’s preference, the GL will maximise the value assigned to the alternatives. Similarly, if A prefers B (B1, B2) to M, we assign a higher numerical value to B (B1, B2) than to M, ensuring that behaviour reflecting preferences maximises the numerical value.

If the GL is indifferent between M and B, it seems reasonable to assert that a rational person would choose M and B (B1, B2) with equal probability of ½ if A is indifferent between the two. By assigning identical numerical values to both M and B (B1, B2), a GL choosing M and B (B1, B2) with equal probability will maximise the numerical value, which is the same for both options due to the underlying indifference. Thus, the sum GL becomes stuck (Sum, due to its irrationality). We have sketched out nothing about a synoptic proof of a representation theorem. We have shown that numerical (growth-utility) values can represent preference relations in a way that preserves the fundamental features of rational use, i.e., one alternative is preferred over the other or they are indifferent.

We can extend this theorem further. Suppose we have three alternatives ai, ej, bk for the GL. If a ? e, and e ? b, where ? denotes ‘at least as preferable as’ or ‘is better than or equal to’, transitivity requires that ai ? bk. In other words, if alternative i is at least as preferable as alternative j, and j is at least as preferable as k, then i must also be at least as preferable as k. For example, if you consider water to be no less preferable than a housing scheme and the housing scheme to be no less preferable than water, then transitivity requires that you must also consider the housing scheme to be no less preferable than water, which negates both social utilities. [I hope this isn't too clear, as it is meant to be somewhat ambiguous]. This transitivity also serves the purpose of rational maximisation because we can honestly state that rational behaviour is simply utility maximisation if the GL’s preferences are complete, transitive, and continuous. In other words, if all the variables (S, B1, B2, A, E, M) continuously maximise utility, they can be termed as utility rationalisable (Aleskerov and Monjardet 2002: 30). In other words, their choices can be justified concerning the underlying utility function.

Let us now define a revolution with expected utility growth. Given variables S, B1, B2, A, E, M, the utility growth u has maximised:

GL … u (S, B1, B2, A, E, M) = u (ai, ej, bk) + · · · + GL u (S, B1, B2, A, E, M)…… (1)

Therefore, there is no definitive conclusion. Be CAUTIOUS.

(Anuvishub Sanjay Tamang is a dilettante economist, Junior Research Fellow, Political Economy,  Centre for Social Innovation and Foreign Policy, Nepal. Views are personal. Email: anuvishub@gmail.com)