Bruce Caldwell has recently argued that falsificationism is just too hard to practice in economics. He takes falsificationism to demand that theories be put to tests which are such that if the results are unfavorable, the theory under test must be given up or substantially modified. But it would obviously be silly to permit the results of any test, no matter how faulty it might be, to lead to such a drastic action. Thus,

Indeed, but if falsificationism really demands all this, it is not understandable why falsificationism is dominant in the methodological literature nor why anybody would stop to consider such an implausible doctrine. For remember that falsifiability is not supposed to be merely some nice trait of theories such as linearity or avoidance of imaginary numbers. In Popper's view falsifiability is the mark of science. And whether something is science or not is not a mere taxonomic question, but a question of whether the claims are genuinely empirically supported or corroborated or not.