Besides, whatever the second round effects (and I for one believe that in a liquidity trap they are better for stimulus than for austerity), austerity clearly starts with a negative first round whereas stimulus' is always positive (hence the name "stimulus").
Pretty hard not to notice that when Government buys a service, a service must be produced.

So the burden of the proof has to be for austerity. And well, we're not even in "results seem positive but there is not enough evidence for statistical siginificance" territory there... Actually, we're even beyond "results appear negative but no statistical significance can emerge yet".

Earth provides enough to satisfy every man's need, but not every man's greed. Gandhi