> It makes no testable predictions That's not true. There are, for example, GDP growth forecasts published by almost all central banks and many other instituti

> It makes no testable predictions That's not true. There are, for example, GDP growth forecasts published by almost all central banks and many other institutions. They are pretty wrong, most of the time. But they are made, and falsifying them has been pretty easy in the past. There are many more theories in economics that are, in principle, testable and falsifiable. The problem is that it's difficult to run controlled "experiments" on interventions. You never have a control group in the strict sense, and everything is connected to everything else.