After last week’s post, I thought that I would simply explore adding De Grauwe’s behavioural functions to the cubic IS model. However, the fact that the former is dynamic while the latter is static sent me on an economic goose chase.
Rather than take De Grauwe’s equations down to the cubic model, perhaps I should be taking the ideas of the cubic model up to more sophisticated frameworks.
So if I was to look for evidence for the cubic IS curve, whether that’s in an econometric or simulated model environment, what should I be looking for:
-For one thing, credit should matter for consumption. This is a long way from the classical view that consumption only depends on wealth, as in the permanent income hypothesis. Beginning with the work of Mankiw, economists have shown that at least some people consume out of current income. There is also evidence that credit matters, as discussed here , so I should see that credit has an impact on consumption.
-Moreover, the essence of the cubic model is that excess credit is distortionary. So I could expect to see that periods of high credit are associated with more volatility than at other times.
-Following on from this, periods of high credit should alter the sensitivity of interest rates and consumption.
Perhaps these last two points lend themselves to an event analysis similar to the research already discussed. Alternatively, they possibly could be explored in the simulated DSGE models like that of De Grauwe’s. Just this week I have come across this macro model database , which may provide models to test these hypotheses.
Consequently, I envision the next 12 months consisting of a convoluted path from my theoretical exploration to econometric and empirical models, and then back again. This process will also allow other ideas to be incorporated. My pseudo research interests lie in two separate but complimentary areas: bringing psychology to economics and adding more financial market imperfections to macro modelling. For example, I’m very intrigued by Akerlof’s views on things such as norms in utility functions, as discussed in this speech .
It will be rather challenging, but I would like to bring those ideas into this process as well.