If the answer were always no (or, conversely, always yes) you'd just write a three-line computer program and save the taxpayers some money. Or, given the voting along fairly partisan lines in Congress and the Senate, maybe it's just the political equivalent of an AOLer "me too!" (I figured that if I ever ran for office, and had some absolutist who spoke in terms of 'always' and 'never' running against me, I'd at least try to get the geek vote by writing a few lines of code and saying the voters ought to just put that in office, rather than my opponent).
George W. Bush didn't feel he needed to listen to
anybody, including his advisors (the Decider in Chief). Barack Obama seems to be at the opposite end of the spectrum--it feels like he doesn't stand his ground nearly enough, in the name of compromise. Now, maybe it's easy for an armchair politician like me to say, given that I don't have to work with the people in question, but it feels like he and the Democrats in general could use a little more resolve.
I think that leadership is about doing the right thing--and let the consequences (usually personal) be as they may. That's a nice talking point, but unfortunately, I think that you have to survive long enough to let your actions come to fruition, and really unfortunately, you've got to have made the 'right' decision (or at least not a completely wrong one). And the 'right thing' short term isn't always the 'right thing' long term, and is often in the mind of the beholder.
You can continually focus group and opinion poll (the Clinton administration did this a lot--but they usually also came out on the right side of the equation), do what you feel like (Bush the younger), or do what you feel like and yet bring the public along to your side of things by force of conviction / charisma (Tony Blair comes to mind). Maybe I'm projecting onto President Obama, but I find analysis and back-of-envelope calculations comparatively simple. Knowing when to cut off the analysis and make a decision, based on what you've gathered--that's hard. Just flip a coin and make a decision, and to hell with the analysis--that's reckless.
As for what constitutes a hard decision, that's probably quite subjective. There's a disputed story that British leadership in World War II knew of an impending German raid on Coventry via intercepts and decoding of German communications. But acting would tip off the Germans to the fact that their communications were being listened in to. So they did nothing, and let people perish and factories get destroyed.
http://en.wikipedia.org/wiki/Coventry_Blitz#Coventry_and_Ultra
Disputed though the account is, that's a hard decision to make. Probably even harder to step back from the brink and not just start thinking of everybody as acceptable casualty. In any case, long story short, yeah, somewhere in the middle--you're elected to make decisions for the good of the voters. But, at the end of the day, you're accountable to them too (and they have an unfortunately short attention span).