The IS/LM model is standard macro textbook stuff. Unfortunately the model is highly problematic and even worse it seems like the IS/LM model (in its most simple form) is the only model that certain policy makers understand. Recently a debate about the IS/LM model has been flaring up.
Tyler Cowen first explains what he thinks is wrong with the IS/LM model.
Then Keynesian blogger and economist Brad DeLong blogs in defense of the IS/LM model.
But DeLong comes under heavy shelling from the Market Monetarist camp:
Scott Sumner, Nick Rowe and David Glasner all weigh in on the debate.
You tell me who is winning the debate…
Kevin Donoghue
/ October 8, 2011Who is winning the debate? John Richard Hicks, I think.
I’ve been wondering how I would explain to a non-economist why people get so steamed up about this. Best I can do: way back in 1937, Hicks set out to make Keynes’s GT accessible to undergraduates trained in traditional neoclassical theory. This enraged two groups of people: (1) leftists, mostly English, who saw him distorting the Gospel according to St Maynard; and (2) conservatives, mostly American, who saw him exposing the pure hearts of their young to commie notions from Cambridge and Bloomsbury. Despite the passage of time, the rage has not subsided.
Hicks, meanwhile, got on with his life. A model is just a model. I’ve been reading his last book, A Market Theory of Money; excellent book. Wouldn’t be a bad name for a blog, eh?
Paul Andrews
/ November 15, 2011If the model is used to justify policy, and the model is wrong, then the policy will be wrong, and may do real damage to real people. A model is not necessarily just a model.
Lars Christensen
/ October 8, 2011Kevin, you might be right – Hicks might be winning even though he of course did not really believe in the IS/LM model himself.
And yes, A Market Theory of Money is an excellent title…
Paul Andrews
/ November 15, 2011De Long: “When you do economics and apply it to the real world, you start with the simplest possible model. Does that help you understand enough of the real world to satisfy you? If not, you complicate it by adding the most important thing that you had left out”
This presumes only one possible simplest model. In fact it is possible to define many models of the same simplicity.
He posits that ISLM is the next order of complexity when we extend from the simplest model. However, each possible model of lowest (order-1) complexity has many possible extensions of order-2 complexity.
Having begged the question by implying ISLM is the obvious next “correct” model of order-2 complexity (when it is merely one choice of many), he claims that we must learn its “insights” before moving to order-3 models.
The problem with this logic is that if the model is flawed, the “insights” may also be flawed. Further, added layers of complexity may completely prove the “insights” to be completely incorrect.
Tyler’s points go to the elements of the model itself, as well as to what the model leaves out.