>BTW, government expenditure only appears in the GDP calculations by expenditure. It doesn't appear in the income nor in the production. The later one the one you are actually thinking of.

I am not sure against what you argue exactly. My point is that if we use GDP as a metric of economy strength, then it's easy for the government to cheat it by fueling it's spending by printed money (ah, sorry, the reputable economist use the "quantitative easing" euphemism instead) or by unfair debt issuing practices (like forcing banks, pension funds, or even people directly to buy government bonds), then it results in a short term "sugar rush" GDP growth, while consequences are kicked down the road. In the US case we also have an additional masking factor in the form of having control over the world's reserve currency, which significantly dampens the usual inflation reaction.

So if we want a better metric, we should somehow discount this factor. Obviously, simply removing the government spending is too simplistic (or you can call it less charitably as "dumb").