Microeconomists and behavioural economists do indeed conduct repeatable experiments. This is more difficult for macroeconomists, of course, but the same is true for many sciences, such as theoretical physics, climate science or much of biology.