[smokeyj]

You missed my point. I'm arguing theorizing around statistical data is only half the battle to the scientific equation. The other part is putting those theories to the test and quantifying the results in meaningful way over time. Again and again. Without the again-and-again part, you only have theoretical knowledge. Theoretical knowledge is not a "hard science". Involving data does not make a hard science. Applying heuristics to the data does not make it a science. A computerized model does not make something science. The weather-man uses science when developing a forecast. The forecast is not a hard science.

Take minimum wage for example. This is a theory with supporters in both camps - some for, some against. Well, how can we ever measure if a minimum wage is having a positive influence? How can we continue in support of a theory with no supporting evidence? Pile on 500 more concurrent theories, and you have a cluster-fuck only Krugman could defend.