Saturday, 8 April 2017

John Broome proving Normative Economics is either strategic or empty



Economist turned philosopher, John Broome, whom I've previously discussed,  inadvertently shows why Normative Economics must be either strategic or empty in an interview with 3 AM

Intentions and Beliefs are opaque to Rationality

3:AM: Starting with reasoning and normativity : if intentions aren’t automatically a reason to do anything then how can intentions engage our rationality like they do?

JB: How does a contract engage the law? Because the law requires you, if you’ve made a contract, to fulfil it; because the law requires you, if you fail to fulfil a contract, to pay compensation; and so on. These are facts about the law and particularly about how contracts figure in it.

How does an intention engage rationality? Because rationality requires you, if you intend an end, to intend whatever you believe is a necessary means to that end; because rationality requires you not to have contradictory intentions; and so on. These are facts about rationality and particularly about how intentions figure in it.

Contracts exist even where there is no Law. If a monopoly of legitimate coercion exists, it may be that the monopolist stipulates that all contracts made or which take effect within its demense come under the purview of a Court or Tribunal or other body. If there is no monopoly, there may still be an oligopoly of some sort and there may be jurisdictional conflict.

However, Contracts don't depend on the Law. They exist so as to clarify the nature of a uncoerced transaction and to widen its scope in a mutually beneficial way. Contracting parties may decide that it is beneficial to recognise a particular Court or other such body as regulating the contract such that a vinculum juris, a bond of law, is created and rights are linked to remedies under different states of the world. There are many reasons why contracting parties may wish to go down this road. In particular, contingencies relating to the death or incapacity of one or other party can be better provided for.

Intentions aren't like Contracts. They are things we refer to when seeking to explain or predict observed behaviour. They may also be things we suspect are our own unconscious motivations. We seek to clarify our intentions and put a check upon them- for example by making vows and submitting ourselves to a moral audit- because we wish to be trusted by others and to trust ourselves.

Intentions create, not Contracts, but Relationships. They increase trust but don't create a viculum juris. However, the operation of law can give rise to a tort- i.e. a civil wrong or negative externality- a defence against which might involve proving one had a laudable intention and sought to fulfil that intention with due diligence. Notice such torts exist whether or not contracts exist. They are created by the Law and wholly independent of Contracts.

It often happens that a conscionable contract is unenforceable in a particular jurisdiction which otherwise recognises its validity. Thus an Islamic marriage is a contract whose sanctity is recognised by many non-Islamic jurisdictions. However, some of the provisions of that contract remain unenforceable.

Thus, Broom is wrong. It is not a fact about the Law that it makes you fulfil a contract because you had the option of not consenting to the Law compelling you to do so at some later date. It is a fact about the Law that it can make you fulfil what it considers the Social Contract requires of you with respect to someone injured by your actions whether or not there was consent or a contract or indeed any sort of relationship between you and that other person.

Similarly it is not a fact about Rationality that it requires you to intend to believe anything at all. Either you intend something or you don't. Either you believe something or you don't. Being required to intend to believe something is silly because no one knows how to reverse engineer either intentions or beliefs.

Suppose I believed myself to be allergic to fatty food. My health would improve because I'd be less fat. I don't know how to engineer this belief. I tried going to a hypnotist. It didn't work. What am I supposed to do? Start doing brain surgery on myself?

If I intend to lose weight and believe that 'the belief that I am allergic to fatty food' is a means to that end, Rationality does not require me to intend to believe that I can gain this belief anymore than it requires me to believe in fairies.

Rationality precludes 'reasons for actions'


3:AM: Philosophers like Nagel and Raz argued that reasons are the keys to normative thinking but you’ve argued that reasons don’t exhaust normativity haven’t you? And should we want to be rational anyway?

JB: Do you mean to ask whether I think normative thinking is concerned with other things besides reasons? Indeed I do. For example, normative thinking is concerned with what we ought to do. In fact, I’d say that ‘What ought I to do?’, ‘What ought I to think?’, ‘What ought I to hope for? and ‘What ought I . . .?’ in general are the core normative questions. Thinking about reasons can help to answer them.

Then you ask whether we should want to be rational. That you ask this question here makes me think you assume that being rational amounts to doing what you have reason to do. I don’t think that. There is much less of a connection between reasons and rationality than many people assume.

Still, the question is important anyway. (Not whether we should want to be rational – I think wants are not philosophically important – but whether we should be rational.) I think it’s not necessarily the case that you should be rational. There are well-know examples, particularly from Derek Parfit, where very bad things would happen if you were rational on a particular occasion. On that occasion, it’s not the case that you should be rational. But I do think you necessarily have at least a reason to be rational, though it could be defeated by stronger reasons not to be rational. I cannot prove this, however.


Rationality is concerned with that which is in accordance with logic and thus has a mathematical expression. A 'reason' defeated by a 'stronger reason' is a mistake that has been discovered, a fallacy that has been exposed. It may still have utility in some applied field as an approximation. But it is no longer part of 'Reason'. Rationality can't require us to have 'reasons' which, if the history of mathematical logic over the last hundred years is any guide, we know will be defeated by stronger reasons. 

It is a separate matter that, for legal reasons, I might decide to quote a particular normative reason as guiding my action- for example in a statement of intentions ancilliary to a Deed of Trust. Here, the reason quoted is an admission of a certain type or process of defeasibility. It allows the Trustee more freedom of action. However, it is not the case that this normative reason needs to be shown to have been binding upon me when I made the Statement of Intentions. In other words, the normative reason isn't normative at all but actually strategic within a particular legal context.

The question of 'ought'- deontics- relates to how to make ourselves more trustworthy even only to ourselves. But this is an empirical matter. I may believe that I 'ought' to lose weight and that what will enable me to achieve this end is believing that a beautiful little fairy will die every time I bite into a custard pie and that the beautiful little fairy's goblin lover will turn up and shoot poisoned arrows into my eyes to avenge her death. However Rationality does not require me to believe in this beautiful little fairy. Why? Fairies don't exist. That is an empirical matter. Similarly, Rationality forbids any requirement felt by anybody to believe in either fairies or Broome's 'reasons'. The thing is silly.

Normative reasons are either strategic or path dependent
3:AM: Do you agree with Nagel when he says that he sees ethics as a branch of psychology? If it is how can armchair a priori reasoning about ethics get anywhere? Don’t we need Josh Knobe’s xphil crew to burn the armchair and work out the empirical experiments to generate useful data?

JB: No I don’t agree with Nagel about that. One of the advances in moral philosophy since Nagel’s The Possibility of Altruism is that we now routinely distinguish normative reasons from motivating reasons. Nagel explicitly declined to do that. Consequently, he did not distinguish the study of normativity from the study of motivation, which is a branch of psychology.

We now make this distinction. Ethics is a part of the study of normativity, and not a part of psychology. It could hardly be independent of psychology, though ,since it is intimately concerned with human minds.


Broome rightly points out that his subject wouldn't exist unless normative reasons are different from motivating reasons. Thus, there is a strategic reason to make that distinction. Hence, at least one normative reason could be the strategic expression of a motivating reason. However, if the reason for upholding the independent existence of normative reasons can be merely strategic, how are we to know if any normative reason we might have is not itself just the strategic expression of a, possibly repugnant, motivating reason? Rationality can't counsel a dubious means to worthy ends if that same Rationality impugns those ends by the same means.

There is a way to circumvent this objection by positing a sort of 'Chinese Wall' or bicameral neurological structure as obtaining in agency. However, this implies path-dependence. But, as Samuelson pointed out, without ergodicity, Economics is empty. Thus, normative economics is either strategic or empty.

No comments:

Post a Comment