It depends on the version of economics espoused.
Classical/Keynesian economics are scientific, and so is Monetarism, since they all draw conclusions, which can be described mathematically, only from observations of transactions of actual human beings.
Neither Marxism and Austrian School "economics" are scientific, since they both depend on deductive reasoning from arbitrarily defined premises, the utter antithesis of a science.
Eh... where is that different from all the classical stuff? The saving grace of the Austrians is that they were at least aware of what they're doing.