"Institutions are organisations or patterns of behaviour built by societies to help solve social or economic problems which the law or private markets cannot fully address."
institutions have existed long before the rule of law, or the existence of private markets.
what he does get right is the fallacy of "ceteris paribus" or "all else remaining equal" that is used in most economic arguments with their limited scope.
Ceteris paribus is a technique for decomposing complex problems into simpler elementary ones, not an assertion about underlying economic realities. It's employed most commonly in microeconomics, which works fabulously, in much the same way as classical mechanics continues to work fabulously for the vast majority of earthbound practical problems. Macroeconomics is beset by problems because our understanding of social and organizations structures is seriously incomplete, in much the same fashion as ants probably have extreme difficulty in forming ideas about the collective behavior of ant colonies.