DSGE, SFC And Behaviour

This is a continuation of my post Simon Wren-Lewis On Wynne Godley’s Models. I was comparing stock-flow coherent models to DSGE models implicitly (didn’t mention the ‘DSGE’).

One of the things I spoke of was behaviour: firms deciding how much to produce. In stock-flow consistent models, it is decided by trends in sales. So if entrepreneurs see a fall in their inventory-to-sales ratio, they’ll produce more typically. This can be made more accurate. See Wynne Godley and Marc Lavoie’s text Monetary Economics for more details.

Here I want to concentrate on models such as DSGE or any other model used by institutions such as the UK Treasury for the case of production. In these models, there is a production function describing how much firms will produce. This is incorrect to begin with. It says nothing about behaviour. If households start borrowing a lot, in DSGE models, producers are still producing the same because production is governed by the production function. In stock-flow consistent models, simple modeling assumptions about how much firms produce are far superior. So in this case, in SFC, more borrowing leads to more sales and a change in sales trends, inventory/sales ratio and hence affecting how much will be produced.

The DSGE production function is thus inconsistent with the Keynesian principle of effective demand. DSGE is not even Keynesian. It’s thus ridiculous how economists defending DSGE models and its ancestors accuse SFC modelers of not paying attention to behaviour.

Leave a Reply

Comments are welcome, but not published—see comments policy. Required fields are marked *