> But there is also some pretty good research that helps you understand the economy.
I'd love to see some. I looked quite hard into two topics, gravity models of trade and the Philips curve, and I never found much I thought was convincing.
I think the culture of academic macro doesn't appreciate just how hard it is to thoroughly test a model. It's not as simple as writing down some dynamics, estimating via maximum likelihood, and drawing conclusions based on the parameters that pop out.
...yep, I mean the gravity model of trade is just a regression. It is very unlike pretty much everything else in macro because its validity is statistical rather than theoretical (and more models that are justified by theory often perform worse...although this is kind of complex).
The Philips curve changes over time. It is, however, quite easily testable so it either applies or doesn't. If you are looking for some hard rule that applies every time then you should take an interest in something else (indeed, this was my initial point...the issue is that economists believe that the economy are a set of equations to be solved...in reality, the nature of the economy is changing constantly).
I don't think that is the case. Economists are confronted with the difficulty of tests everywhere (the issue is that they often don't handle this well), it is why econometrics is distinct from statistics. As an example, interest rates and growth are positively correlated (i.e. higher rates appear to cause higher growth)...we know this isn't the "true" relationship (interest rates are a function of expected growth) but there has been a ton of work done to separate these effects (this isn't only in macro, controlling for confounders/ommitted variables is really what econometrics is about). Econometrics is really the best part of economics.
> interest rates are a function of expected growth
Again, you're saying that as if it's a matter of fact. When I looked into the relationship between growth and interest rates, I found baffling discussions of demand for money as it relates to a financial system completely unlike the one we use today.
I just find it impossible to separate the nonsense from the useful insight. At some point I just gave up.
It is a matter of fact because interest rates are determined by the Fed...who say: we look at expected growth. Don't make things more difficult than they need to be (you can also see this quantitatively).
Short rates are determined by the Fed. And they target inflation, not growth. The Fed has relatively little control of the yield curve at long maturities.
I'd love to see some. I looked quite hard into two topics, gravity models of trade and the Philips curve, and I never found much I thought was convincing.
I think the culture of academic macro doesn't appreciate just how hard it is to thoroughly test a model. It's not as simple as writing down some dynamics, estimating via maximum likelihood, and drawing conclusions based on the parameters that pop out.