During QE the central bank can also acquire other assets that are riskier than bonds and take risk off other people's balance sheet, again hoping that money will be used to stimulate the real economy.
At 0% it pretty much stops working. Up to 0% some bond holders are still holding on because they are betting on that being the better investment. The other problem is how do you exit QE. We're going to find out how that works over the next few years...
But at a certain point, interest rates are low enough that lending is limited by the lack of creditworthy borrowers or by bank capital requirements, and so further reductions will have very limited impact.
You are correct - at a certain point there are no investments with a positive real rate of return. At this point all further resources should be devoted to consumption regardless of time preferences (interest rates). If consumption desires are also completely satisfied, then you are in a post-scarcity economy.
Almost by definition [1], when you reach that point, you are NOT in a recession.
What's the relevance?
[1] The technical definition of a recession is 2 quarters of negative growth, but for theoretical purposes it generally means a period when real resources are not fully utilized.
At a certain point there are no investments with a positive real rate of return. If consumption desires are also completely satisfied, then you are in a post-scarcity economy. Almost by definition [1], when you reach that point, you are NOT in a recession.
You don't have to be in a post-scarcity economy to have no lending opportunities at the margin. It is also very possible to have no additional lending opportunities, at any positive interest rate during a recession. This is especially so with the way that capital requirements work. If a bank doesn't have adequate capital ratios, they cannot lend, no matter what the rates.
The salary loss aversion I'm talking about is taking a 10% pay cut in a deflationary environment vs. not getting a raise in a zero inflation environment or getting a raise smaller than inflation in an inflationary environment.
The pay cut is going to feel much worse. That's loss aversion. From my exposure to behavioral economics through Dan Arieli I'd say the real loss of $1 is going to feel a lot worse than the nominal loss of $1 eroded by inflation. The latter feels like nothing which is why governments can use inflation as a tool so effectively. The imagery of high inflation is of people rushing with their bags of money to spend it before it becomes worthless... I've lived through some periods of high inflation and when interest rates were at over 10% I don't recall people rushing to invest, they rush to spend...
http://en.wikipedia.org/wiki/Nominal_income_target
From a behavioral/psychological perspective, interest rates seem more likely to affect consumption choices. All of Keynesian and monetarist economics is predicated on the idea that people are ignorant of real prices and focus on nominal prices. So 2% nominal interest - 0.5% inflation looks like a 2% rate of return (invest more!) whereas 0% interest + 0.5% deflation looks like a 0% rate of return (consume more!).
[edit: the salary example is NOT loss aversion. If you lose your job for $100, and turn down a job paying $90, you "lose" $100. If you accept the job you "lose" $10.]