Matt Nolan

Endogeneity is one of the advantages of economics above naive heuristics

/in /by

Sometimes you hear a comment that helps you see that something may be unclear, when economists may not have thought it was previous:

(ENPOG here stands for endogenous potential output gap)

The statement in that tweet made little real sense to me [Note, I could build an explanation – say that the endogenous credit cycles he’s saying have some impact on measures of potential output, so that you get a non-inflationary cycle. Of course, this doesn’t say too much – unless you are willing to take it as far as Borio. But remember with all this we are talking mainstream economics again …].  In the same way that I found it weird how people went on and on about endogenous money/credit as a way of seemingly winning a myriad of entirely different arguments!

The key thing is that thinking about issues being “endogenous” is essentially what economic methods do.  Supply and demand are a quintessential example of a model with endogeneity – price is endogenous in this model.

The best way to think of endogenous is that we are saying something (a variable) is determined “within the model”.  Single models are often partial, so they involve explaining some endogenous variables keeping other exogenous (outside the model).  This simplifying assumption is sometimes shown to be inappropriate (Lucas Critique) and as a result, we try to use “many models” in order to balance understanding.

Dealing with endogeneity in our explanations and models is something economists are totally and utterly obsessed with.

The thing that ties all economists together, both mainstream and “heterodox” appears to be a view that assumptions about expectations matter.  When the other tweet mentions Steve Keen it reminds me that the endogenous credit cycles he builds needs to be based on some underlying choice of agents, and that appears to be based on a certain view of expectations.  Since the Lucas Critique (and to be far in large parts prior) mainstream economists knew that, unless they actually explicitly accounted for expectations, we do not really know how policy will influence outcomes – we don’t have enough “structure”.

Naive heuristics are often the only replacement for the ideas set down by economists.  The obsessions of folk economists can lead them to make systematic errors, and as Rubin states one of the key focuses of economists should be on pointing out these systematic errors – having an idea of underlying structure, and looking at social structures in a more complicated way then uni-causal chains … and this is what economics, and the sometimes maligned general equilibrium theory, does.

Endogenous is not a catch phrase we can throw around to critique things we don’t like – (although I do love whipping it out when talking about econometric models 😉 ).  We have to actually state what simplifying assumption/missing element is important for driving the difference we have, and then state that.  This is how the economic method (both theoretical economics and econometrics) is useful 🙂 .