Nearly every forecast now says that, in the absence of strong policy action, real GDP will fall far below potential output in the near future
we need a fiscal stimulus big enough to close a 7% output gap
These brief statements highlight a major problem with Keynesian economics, the assumption that there is one ideal "potential output", or worse, that the ideal level of output is that which is the greatest. What this fails to consider is that output based on debt is generally not good. To put it another way, growth based on debt and speculation is false growth. As we've seen recently, false economic growth eventually reverts, which can lead to quite severe crashes. So this leaves us with some questions: What sort of growth should we look to achieve? Is perpetual growth by any means possible the goal? Is contraction always bad, and if not, is there some magical level beyond which contraction becomes bad?
In the Keynesian view of the economy where consumer spending drives growth, merely stimulating consumer spending is the theoretical key to prosperity. But if this were so, why not take on unlimited amounts of debt, stimulating unlimited amounts of consumer spending in good times and bad? The fact that such a scenario would lead to disaster is not lost on Keynesians, which is why they instead look to achieve optimal growth, and stimulate spending if needed via the optimal amount of debt, whatever those things may be. If consumption can be stimulated just enough to increase output, but not so much that the cost of the stimulus outweighs the increase in production, then we've achieved something good. But how can we measure the result of our stimulus to know if it's been successful? How do we know if we've not simply created more false economic growth?
Keeping in mind that increased production based on consumption is the goal of fiscal stimulus, the success or failure of stimulus would clearly lie in what is or isn't consumed. This brings us to an important distinction, productive consumption versus nonproductive consumption. With productive consumption, a participant consumes based on what they produce. Their production not only enables them to maintain their well-being through consumption, their consumption sustains their ability to produce. Obviously not all consumption is productive though. While some economic participants are able to produce much more than they need to consume, which is clearly productive, other participants consume much more than they are able to produce, or they produce goods which are not useful for sustaining further economic production - this is clearly nonproductive. If the goal of fiscal stimulus is to grow production, we would certainly want to support the economic activities that produce the most useful production sustaining goods with the least amount of input. But herein lies the problem with stimulus policy, it makes no distinction between productive and nonproductive consumption, nor can it.
So what happens when we implement a fiscal stimulus? Realistically, it is impossible to measure and distinguish those economic activities which are productive from those which are nonproductive. Therefore, the result of stimulating consumption is a blanket subsidizing of both productive and nonproductive economic activities. Unfortunately, subsidizing those economic activities which are nonproductive results in false economic growth, which as always, will painfully revert at some point. Looking at recent history, we can clearly see that the growth of the housing and financial services sectors were beyond what was needed to sustain other economic participants. While their growth was measured as increased output, clearly it was false growth and has since receded.
Getting back to the original topic of implementing stimulus in order to achieve "potential output", two things are now clear: some portion of growth resulting from stimulated consumption will inevitably be false, and stimulating nonproductive activities is not only counter-productive, it is harmful to productive activities that are competing for the same resources.
Ultimately though, all that anyone is cares about is whether or not the end will justify the means. Do the benefits of fiscal stimulus outweigh the costs? Can we achieve potential output through fiscal stimulus, and more importantly, do we even know what an optimal output is? The problem with the calls for fiscal stimulus isn't the idea of stimulus itself, but with the fact that such calls fail to address the many inherent problems and questions raised here. For all of the very real costs that stimulus carries, we should hope for more from its proponents including Krugman. Anything less is shooting in the dark.