“The fact that no responsible government would ever literally drop money from the sky should not prevent us from exploring the logic of Friedman’s thought experiment, which was designed to show—in admittedly extreme terms—why governments should never have to give in to deflation.”
The quote above is from a post by Ben Bernanke (who, in case anyone does not know, used to be in charge of US monetary policy). I put it up front because it expresses a macroeconomic truth that no one should ever forget: persistent recessions and deflation are never inevitable, and always represent the failure of policy makers to do the right thing.
There are many useful points in his post, but I just want to talk about one: Bernanke is in fact not talking about helicopter money in its traditional sense, but what I have called elsewhere ‘democratic helicopter money’.
When most people talk about HM, they imagine some scheme whereby the central bank sends ‘everyone’ a cheque in the post, or transmits some money to each individual some other way. It is what economists would call a reverse lump sum tax, or reverse poll tax: the amount you get is independent of your income. That makes it different from a normal tax cut.
In practice the central bank could only really do this with the cooperation of governments. It would not want to take the decision about what everyone means on its own. (Do we include children or not. How do we find everyone?) But once those details had been sorted out, a system would be in place that the central bank could operate whenever it needed to.
Bernanke suggests an alternative. The central bank sets aside a sum of newly created money, and the fiscal authorities then spend it as they wish. They could decide to use all the money to build bridges or schools rather than give it to individuals. There might be two reasons for doing HM this way. First, for some reason the fiscal authorities are reluctant to spend if they have to fund it by creating more debt, so it may allow them to get around this (normally self-imposed) ‘constraint’. Second, a money financed fiscal expansion could be more expansionary than a bond financed fiscal expansion. Lets leave the second advantage to one side, as the first is sufficient in a world obsessed by government debt.
I have talked about something similar in the past (first here, but later here and here), which I have called democratic helicopter money. This label also seems appropriate for Bernanke’s scheme, because the elected government decides on the form of fiscal expansion. The difference between what I had discussed earlier under this label and Bernanke’s suggestion is that in my scheme the fiscal authorities and the central bank talk to each other before deciding on how much money to create and what it will be spent on (although the initiative always comes from the central bank, and would only happen in a recession where interest rates were at their lower bound). The reason I think talking would be preferable is simply that it helps the central bank decide how much money it needs to create. 
Imagine, for example, you had a fiscal authority in one country that wanted to spend the money on ‘shovel ready’ public investment projects, and an authority in another country that wanted to spend it on some temporary tax cuts for the rich. The impact of the two different stimulus policies on demand and output are very different. If the two economies were in similar conjunctural positions, then the central bank with the tax cutting fiscal authorities would want to create a lot more money than would be required in the other economy.
In some countries it is easier for central banks to talk to the fiscal authorities than in others. When it is difficult, Bernanke’s scheme may appear attractive, but it leaves the central bank somewhat in the dark about how much money it needs to create. The big advantage of the more popular conception of HM (a cheque in the post) is that the impact of any money creation is much clearer. (As it is important to end recessions quickly, waiting to see what happens is not helpful advice.)
When central banks and governments do happily talk to each other (as in the UK, for example) then my version of democratic HM becomes an option. Arguments that this makes the central bank less independent are spurious in my view. The central bank initiates the discussion, in clearly defined circumstances. They simply ask what the government would spend any newly created money on. This question should be accompanied by the central bank’s current view on what the multipliers for various fiscal options are. The government then makes a choice, and the central bank then decides how much money to create.
While democratic HM is not talked about much among economists (Bernanke excepted), I think there are good political economy reasons why it may be the form of HM that is eventually tried. As I have said, conventional HM of the cheque in the post kind almost certainly requires the involvement of government. Once governments realise what is going on, they may naturally think why set up something new when they could decide how the money is spent themselves in a more traditional manner. Democratic HM is essentially a method of doing a money financed fiscal expansion in a world of independent central banks.
Which brings me back to the quote at the head of this post. The straight macroeconomics of most versions of HM is clear: all the discussion is about institutional and distributional details. If it is beyond us to manage to set in place any of them before the next recession that would be a huge indictment of our collective imagination, and is probably a testament to the power of imaginary fears and taboos created in very different circumstances.
 A sequential set-up of the kind Bernanke suggests is also more vulnerable to cheating: the government uses the money to finance something they were going to do anyway, and in effect largely offsets the money creation by reducing its own borrowing.