While I'm sympathetic to the goals of empirical studies of this type, I doubt their utility: if they find that rent control increases rents, the evidence is rig

While I'm sympathetic to the goals of empirical studies of this type, I doubt their utility: if they find that rent control increases rents, the evidence is right; if they find that rent control does not increase rents, the evidence is wrong. The reasons for this apparent rejection of empiricism are simple: (1) Empiricism, so successful in the exact sciences, relies on clean evidence and controlled experiments, but evidence in economics is necessarily messy, and controlled experiments rare. (2) Given extremely reasonable and general assumptions (buyers tend to buy more of something if its price goes down, producers tend to produce more of something if its price goes up), the connection between rent control and rents is a theorem, not a conjecture. (See any introductory economics text for a proof.) Doing an empirical study to test the statement "Rent controls increase rents" is kind of like measuring the length of ladders leaning against buildings to test the Pythagorean Theorem. You're likely to find that Pythagoras was right, but if you don't it's probably a flaw in your measurements rather than a mistake in the proof.