James wonders why historians of thought who are also textbook writers (I
wonder who he has in mind)  have not put this "theoretical fraud" to
rest. The issues are complicated. First, IS/LM, while it has problems is
not a theoretical fraud, and its history and derivation has been
extensively studied. Back in the 1960s it really did form the way
economists thought of macro modeling--I agree by obscuring dynamics--but
at the time, they didn't have the modeling capabilities to handle
dynamics. They do now, which is why IS/LM and AS/AD are not really
discussed in modern macro theory.

Why does IS/LM persist? I've written about this in "The Strange
Persistence of IS/LM" in HOPE which is reprinted in my collection of
essays "The Stories Economists Tell". Essentially, my argument is that
it serves as a graph to hang discussions of policy on, and that's what
the focus is in the undergraduate macro theory course.  Interestingly,
Ackley's 1960 macro theory book (which was meant to teach macro theory,
not macro policy) had only 3 IS/LM diagrams, and he called it a minor
expositional device. Mankiw's book (which is meant to teach policy) has
27 IS/LM diagrams, and it is the central expositional device.

In my texts I have tried to make clear what the issues are, but when I
have developed them in full--reviewers said the presentations became too
complicated for students. They are not interested in theory--just
policy. I think it is important to see textbooks for what they are
pedagogical devices to convey ideas to students.  They often differ from
the way economists who are working in the subject see the issues.

Dave Colander