Arrogance and self–satisfaction among macroeconomists

Simon Wren-Lewis (Crisis, what crisis? Arrogance and self-satisfaction among macroeconomists) blogs that there is nothing fundamentally wrong with modern micro-based, so-called New Keynesian, macroeconomic models despite their evident failure during the financial crisis:

Let me be absolutely clear that I am not saying that macroeconomics has nothing to learn from the financial crisis. What I am suggesting is that when those lessons have been learnt, the basics of the macroeconomics we teach will still be there. .....However, this must surely use the intertemporal optimising framework that is the heart of modern macro.

The problem is in the last sentence. The use of a microeconomic optimising framework that he insists on severely limits the scientific value of the model. We cannot do meaningful macroeconomics, which is all about income, employment and the price level, using microeconomic tools.

First, it is surely unreasonable to model the behaviour of a single representative household as standing for a whole society of millions of individuals. That shifts the analysis away from the aggregate behaviour of the whole economy that we are interested in understanding. Instead of focussing on the problem of effective demand, which is all about the circular flow of spending among income-constrained households and firms, it nets out all that is important.

But, perhaps more importantly, it fails the fundamental test of a scientific model: that it must faithfully reflect the real world it is describing. There are strong grounds for believing that people, or a substantial number of them, are not doing the kind of continuous intertemporal optimisation at the centre of the model.

Compelling evidence for this is in the latest Scottish Widows Saving and Investment Report.

It finds overall retirement savings have hit an all-time low, with only 46 percent of working people in the UK saving enough for their retirement. It reveals stark differences between people's expectations and reality.

Despite a big fall in retirement savings, it seems that people's aspirations for their pension income have actually increased by £200 from 2011 to 2012. The survey findings show that the average level of annual income people would feel comfortable living on at 70 years-old is now £24,500 compared to £24,300 in 2011. But the savings rates reported are unlikely to be able to produce a pension at age 65 of more than half that figure. Moreover 41 percent say they would like to retire at age 60.

That is not all. More than one-in-five (22%) of people in work aged over 30 and earning more that £10k are saving nothing at all for their retirement.

This is evidence of great heterogeneity of behaviour, with some households optimising like the model says, but as many failing to do so. But what it does show clearly is that intertemporal optimisation is not at the heart of the real economy.

How, then, can we put any faith in an economic model that assumes everything to be driven by universal intertemporal optimisation of consumption by a single representative consumer?

Surely a truly scientific model that would have some chance of successfully tracking the real world would try to model the actual behaviour of real people. That means a statistical approach that captures behaviour and allows for heterogeneity - "by and large and on the average" was a phrase used by old Keynesians for this kind of behavioural economics.

Dennis Leech : 13 Jul 2012 16:08 |  Tags: Keynesian Economics Macroeconomics Wren-Lewis |  Comments (0) |  Report a problem