Continuing the recent theme of heterogeneous households and macroeconomic models (see here & here), I have come across some very interesting research from the field of Social Psychology. In the last post comparing the models of De Grauwe and Woodford, I finished with the view that I had to try and alter the micro-foundations of the Woodford model from the beginning, rather than try to impose some De Grauwe parameters midway through the model derivation. What I would like to attempt now is to work through the Woodford model, but take De Grauwe’s concept of household heterogeneity – that is, instead of borrowers and savers in the Woodford model, distinguish households by fundamentalists and extrapolators.
One of the first issues that I needed to look into was related to Blanchard’s query about the Woodford model. Blanchard questioned whether it made sense that borrowers and savers had different marginal utility of consumption curves. It seems a fair point, but as I would like to use the De Grauwe distinction, could I find some evidence from the Psychology literature about optimism and its effects on discount rates that would enable this distinction. In other words, if there is evidence that optimistic individuals display a different rate of time preference to non-optimists, we might be more comfortable to make the claim that the marginal utility of fundamentalists and extrapolators really are different.
Luckily, I came across a 2001 paper from Acta Psychologica, from Mariette Berndsen and Joop van der Pligt (hereafter B&P). To the best of their knowledge, these authors were the first to have a detailed look into the role of optimism and intertemporal choice. They began their exploration of this topic by looking into Discounted Utility Theory, of Fishburn & Rubinstein 1982, and Samuelson 1937. This theory relies on the notion of present value, which is the current value of a future payment discounted in a way that makes it equivalent to some value in today’s terms. This is shown in the formula below, where Vp is the present value, Vd is the value of the future payment and d is the time period of delay.
In this formula, r is the annual discount rate. Economists view this as the interest rate that the future payment is discounted by, which is simply the reverse of the compound interest rate that grows a present value over each period. In this psychological study, however, it is viewed more as a ‘discount’ on the future, rather than an interest rate per se.
In psychological terms, the discount rate, or rate of time preference, reflects the preference for an immediate gain over a future gain. It is in this framework that B&P explored the possibility that optimists have a different discount rate to non-optimists.
What they found
B&P did found that optimists had different discount rates to non-optimists. They showed that time preferences decreased when optimism of obtaining additional gains were reduced. The authors placed this effect of optimism in the context of uncertainty about the future. It is this uncertainty, they argued, that allows optimism of future additional gains to lead to strong preferences for immediate gains.
How they did it
Participants in this study were 83 students at the University of Amsterdam, who were randomly assigned to one of four experimental conditions. These consisted of 2 versions of outcomes, either gain or loss, and two variants of optimism, high or low. In the gain scenario, participants were working for a company that was doing well and would be paying a series of bonus payments in the future. The optimism regarding the potential for additional gains was lowered for some participants. This was achieved by articulating measures such as adjusting the payments for inflation, locking the participant into a 10 year contract (meaning they could not think of taking the bonus and moving to a higher paying job) and confirming that the bonus payments would be provided only once.
What it means for the macro model
The findings of this paper make a lot of sense to me. I can see how increased optimism about the future and additional gains therein will increase the preference for immediate gains. In terms of the modelling process that I am attempting to undergo, this provides a sound building block to derive differing utility functions for optimists and non-optimists – or in De Grauwe’s terminology, for extrapolators and fundamentalists. Now I need to return to the early stages of Woodfords model to rework the heterogeneous households with these time preferences in mind.
 Berndsen, M, & van der Pligt, J, (2001). Time is on my side: Optimism and Intertemporal Choice. Acta Psychologica108, 173-186.