It wouldn't necessarily bolster the case for stimulus to assume that people will not correctly estimate the costs--people could overshoot as well as undershoot, meaning that they'd actually oversave to pay for future taxes. What matters is, first, are people paying more or less attention to future taxes than they used to, and second, are their estimates more or less optimistic than they used to be?
It's armchair sociology, of course, but I'd argue that people have suddenly become much more focused on estimating their future income and expenses, rather than living paycheck to paycheck--hence the suddenly renewed interest in savings. They've also, empirically, become a lot more pessimistic--at least, if we can count consumer confidence indices as empirical evidence. That will impact how much of the stimulus they save. This does not mean that there will be no multiplier--only that it will be lower than in an era less future focused.
This article available online at:
http://www.theatlantic.com/business/archive/2009/02/how-forward-looking-are-we/252/
