If you are a New Keynesian sticky price kind of person: the answer is also simple. Let the Fed do its magic by lowering the interest rate to stimulate aggregate demand. I have a different answer: replace long dated Treasury bonds in the hands of the public with cash or with short dated Treasury bills.
Many Keynesian economists, journalists and bloggers have argued that, when at the zero lower bound (ZLB) we must repair our infrastructure. Build roads. Build bridges. Build airports. They argue that, when the overnight rate is zero and the thirty year rate is lower than it has been for a century, public infrastructure should be paid for by borrowing at the long end of the yield curve. Float thirty year bonds. Better still: issue Consols that will never be retired.
While I agree that public expenditure in a depression may be helpful: issuing long bonds is not the right way to do it. I agree with Adair Turner that it is better to finance an expansion by printing money or borrowing in the Treasury bill market. Better still: as I argued in How the Economy Works and as Mark Blyth and Eric Lonergan have argued (here) print money and give it to those who know how to spend it: that would be you and me.
Borrowing at the long end of the yield curve is a bad idea because there are still active private participants in that market. There is not one interest rate: there are many. And although it is not possible to crowd out private expenditure at the short end of the yield curve; it is still possible to crowd out private expenditure at the long end.
The maturity structure of debt in the hands of the public matters. As I argue here, it matters because our children and our grandchildren cannot participate in financial markets that open before they are born.
Once one recognizes that the way that public expenditure is financed matters: it is a short step to recognizing that it is all that matters. If the Treasury increases the stock of thirty year bonds in the hands of the public, it will drive up long yields and crowd out private expenditure. If the Treasury reduces the stock of thirty year bonds held by the public, it will lower long yields and crowd in private expenditure. That leads to the argument for Qualitative Easing. A policy that removes long bonds (or other long dated risky securities) from the hands of the public and replaces them with cash or with Treasury bills, will crowd in private expenditure and increase aggregate demand.
Critics of QE have argued that QE3 was less effective than QE2 and QE1. That is true. But Fed intervention in the asset markets was undone by the Treasury that was simultaneously changing the yield composition of its debt to take advantage of low long-term interest rates. My message to the Fed and the Treasury is simple: can we please play cooperatively? Much more coming soon on this topic in a forthcoming book.