Dr. Alexander Douglas specialises in the history of philosophy and the philosophy of economics. He is a faculty member at the University of St. Andrews in the School of Philosophical, Anthropological and Film Studies. In this series, we discuss the philosophy of economics, its evolution, and how the discipline of economics should move forward in a world with increasing inequality so that it is more attuned to democracy. Previous sessions can be found here in part 1, part 2, part 3, part 4, part 5, and part 6.
Scott Jacobsen: With psychology classified as a natural science by you, what are the most substantiated and broad-reaching strong conclusions of psychology relevant to economics?
Dr. Alexander Douglas: I’m no expert on this. Behavioural economics is the main area in which the findings of clinical psychology have been integrated. The major challenge attacks, as Robert Sugden puts it, the notion of ‘integrated’ preferences, according to which each agent is defined by a stable set of preferences that has to be tailored to fit her choice behaviour in all circumstances. So if I choose soup over salad today, and salad over soup tomorrow, then the assumption that I am rational compels us to redefine the objects in my preference-set. It would be irrational to prefer salad to soup and soup to salad tout court, but not, e.g., to prefer soup to salad when I’ve eaten 1000 soups in my life but salad to soup when I’ve eaten 1001 soups.
But is it rational for what I’ve eaten in the past to influence what I choose today? What about the lighting in the restaurant? What about what other people are eating? And then, of course, every soup is unique and every salad is unique: perhaps I prefer this soup to this salad, but not that soup to that salad. But then if the descriptions under which I choose become so specific, economic predictions become impossible: nothing about what I choose today will inform us about what I’ll choose tomorrow, since tomorrow everything will be slightly different.
Economists, it turns out, make a lot of implicit assumptions about what can and what can’t go rationally into what is called the ‘framing’ of a choice: past consumption is permitted to be relevant, but not seemingly extraneous factors like the day of the week on which a choice is made. But who is to say what it is rational to consider relevant to a choice? A lot of behavioural economics is about coming to terms with the importance of framing; people can be found, e.g., to choose to save 98 out of 100 lives but not to condemn two out of 100 people to death. Behavioural economics seeks to know how people typically frame their choices, and how the framing affects what they choose.
In a way, it tries to honour the ideal of ‘value-neutrality’ that underpins modern economics: it looks like a value-judgment to say that past consumption can rationally influence a choice but not the day of the week. Behavioural economists want to get by without even that value judgment. We shouldn’t say that people are irrational just because they take to be relevant what economic theorists take to be irrelevant.
Sugden believes, by the way, that even without identifying people’s preferences as such we can make some judgments about the sorts of economic institutions that they would rationally choose. I’m sceptical. He believes that people will rationally choose an economically liberal arrangement, in which free agents can engage in voluntary exchange in pursuit of a better allocation to themselves – and so they might, under that description. But how about under the sort of description Thomas Carlyle might give to such an arrangement: an unearthly ballet of higgling and haggling, conducted by little profit-and-loss philosophers; an array of pig-troughs where the pigs run across each other in unresting search of the tastiest slops, etc. etc.? Framing matters when agents ‘rationally’ choose institutions, just as much as when they ‘rationally’ choose goods. Public choice theory, I think, must also come to terms with the centrality of framing.
Jacobsen: How might, or are, these most substantiated and broad-reaching strong conclusions of psychology influence the philosophizing about economics?
Douglas: Once we bring framing into the question, I think the whole way of modeling human behaviour has to radically change. I don’t see how this can be avoided. A standard ‘utility function’ in economics will look something like this: U=f(x), where U is the overall utility or wellbeing of an agent and x is some vector of magnitudes, each representing the amount of a certain good consumed. To take framing into account, we’d need to replace x with a vector of descriptions of goods. These can’t be simple magnitudes, and so the whole project of a mathematisation of human behaviour is undermined. Could you not just expand the vector of magnitudes to have one argument for every good consumed under every possible description? You’d have one magnitude for coffee in the morning on my own, one for tea in the afternoon with a friend, one for tea in the afternoon with a work colleague, one for coffee in the evening with my beloved, etc. etc. The problem, of course, is that every good will fall under an infinite number of possible descriptions. And worse, there are descriptions of descriptions: choosing off a menu isn’t the same as choosing from a buffet, and so on.
Moreover, it is hard to see how we can get solid experimental evidence on how people frame choices. We might, using the above example, find that people will choose to accept the loss of two people but not to condemn two people to death. These framing effects matter a great deal, as our spin doctors know well. But how do we define the difference? That too is far from clear – our spin doctors know that too. I think that properly taking these subtleties into account would make economics into a qualitative, hermeneutic, ‘soft’ science – more akin to anthropology than physics.
Behavioural economists are attempting to walk the tightrope between hermeneutic anthropology and quantitative science, but I believe that the tightrope is of infinitesimal width, and sooner or later they’ll topple over onto one side.
Jacobsen: Do any of the aforementioned strong conclusions influence the treatment of time-inconsistency first considered by Spinoza and into the present with professional philosophers such as yourself?
Douglas: Spinoza has an idea of rationality that, I think, sits very badly with economics in general. For him it is irrational to discount the future at all. I might prefer one marshmallow today to two marshmallows tomorrow, but tomorrow I would, if I could, certainly not give up two marshmallows to have had one in the past. It is arbitrary to identify myself with myself at a particular moment in time. Thus he says that the rational person does not value a good differently depending on whether it is past, present, or future (Ethics 4p62).
When modern economists talk about time inconsistency, they mean something much weaker than this. They’re talking about a time-discounting function that is hyberbolic, or generally non-linear. Only a few concede that time-discounting, in general, is irrational; Joan Robinson calls it ‘an irrational or weak-minded failure to value the future consumption now at what its true worth … will turn out to be’ (The Accumulation of Capital, 394).
If agents didn’t engage in time-discounting, economic explanations of interest rate, profit, and so on wouldn’t work. Economists certainly don’t want to say that economic equilibrium depends on profound irrationality in the agents involved. In fact, I think you could argue that their equilibriums depend on forced labour or coercive extraction of some sort. If I take on a loan today, my future self will have to work to pay the interest. He gets no direct benefit from what happened in the past. Or, even if he does, he is unlikely to set the relative value of the past benefit as high as his past self did. But he simply wasn’t consulted in the decision. My past self can be paternalistic or exploitative towards my future selves, but, in any case, there is a dictatorship of the present. Economists treat as coercive a situation in which the preferences of a select group determine the outcomes for everyone. But that is exactly what happens when, in their models, agents at time zero determine what all their future selves will pay and receive, by negotiating with other agents present at time zero.
We could, of course, identify all the future selves of an agent with that agent at time zero, but then we would have an agent with deeply inconsistent preferences. Again: today I prefer to give up the promise of two marshmallows tomorrow for one today, but tomorrow I certainly wouldn’t give up two marshmallows in order to have had one in the past. So a single diachronic agent with a nonzero time-discounting rate would have preferences that are not just ‘inconsistent’ in some weak sense but plainly contradictory.
This isn’t only an academic exercise; it gets to the heart of why markets can’t plan – an issue rendered very palpable in our day by the climate crisis. James Galbraith points this out somewhere in The Predator State. You shouldn’t make the mistake of thinking that futures markets allow markets to plan: what they allow is for present agents to divide up the spoils of what they plunder from future generations by contractual obligations or irreversible natural processes. In this way, as in many others, Spinoza has never been more relevant.
Jacobsen: Thank you for the opportunity and your time, Alex.
Photo by Luis Quintero on Unsplash