The need for empirical support is an interesting one. What does it mean for the empirical data to support a hypothesis? A good theory might not have empirical support if the data does not exist to test it. However, this question brings to mind a related topic.
Robert Lucas (University of Chicago) is famous for something called the Lucas Critique. Essentially, it is naive to try to predict the effects of a change in policy entirely on the basis of relationships observed in historical data.
A good example is the observation that, in American Football, there is too much kicking on 4th down. Thus, we should eliminate 4th down to reduce kicking.
In other words, empirical support needs economic theories.