Who makes sure the long run consolidated government-central bank budget constraint balances? Is it the government? Or is it the central bank? The former gives us a Ricardian regime; the latter gives us fiscal dominance.
Most macroeconomists take a Ricardian regime as their unspoken assumption; and, if pressed, would recommend a Ricardian regime, as is required, in the long run, by central bank independence. If the central bank is to have the independence to pursue its own objectives (be it inflation targeting or whatever), the government must take long run responsibility for its own deficits, grateful for whatever profits from seigniorage revenues the central bank deigns to hand over. Worries about the "coordination of monetary and fiscal policies", heard often in Canada until the mid-1990's, were code for the desire for a Ricardian regime.
Only the pathological cases, the Zimbabwes of monetary policy, have fiscal dominance. When the soldiers need paid, and you cannot or will not raise taxes to pay them, and you cannot borrow because prospective lenders know the soldiers will still need paid tomorrow, and you won't be willing or able to raise taxes tomorrow either, you have to print money to pay them (because they are soldiers, after all), and you might as well print to pay interest on existing debt as well, while you have the paper and ink. You need never default on debts that promise to pay only what you yourself can print.
But the very helicopter money that could help us escape the current recession assumes fiscal dominance. By literally giving away money, the central bank is destroying the assets=liabilities identity of its own balance sheet. It's trashing its own net worth. Call it a money-financed transfer payment if you wish to respect proprieties, or a money-financed tax cut. But the point is the same. The seigniorage isn't handed over to the government; it goes to the people underneath the helicopter. The time-paths of future government spending and taxation stay the same. They must, if the increase in the money supply is to be permanent. And it's that very permanence, with it's implied increase in the expected future price level, that creates the higher expected rate of inflation, and/or higher expected real growth rate, which is what rescues us from the liquidity trap.
Under a Ricardian regime, an increased primary deficit today would need to be offset by an increased primary surplus sometime in the future, with an equal present value. Under fiscal dominance, a temporary increase in the primary deficit today would have no effect on future primary deficits or surpluses; the money-finance would be permanent. That would make the fiscal-cum-monetary policy multiplier much larger. It is equivalent to promising a permanent increase in the long run equilibrium price level. It is equivalent to promising to keep monetary policy loose for too long, even after the recession has passed.
Right now, a little bit of fiscal dominance is just what the doctor would order. Will it be inflationary? Of course! At least, we sure hope people believe it will be. That's the point. If the counterfactual conditional is deflation, we want a policy that delivers an increase in the actual and expected rate of inflation.
Helicopter money, that monetarist thought-experiment, assumes fiscal dominance. So does the fiscal theory of the price level, with its implicit assumption that the present value of primary surpluses is exogenous with respect to the existing level of debt. So does Neo-chartalism (at least, the latter makes more sense to me if I assume that they assume fiscal dominance, which is what I mean by an "implicit assumption").
Fiscal dominance is not an accounting statement. It is not even an economic statement. It is a political (or, at least, political-economic) statement about how the world works, or, more precisely, how people expect the world to work. If push comes to shove, and the government's tax, spend, and debt-service objectives conflict with the central bank's monetary policy objective, who has the last word? Or, rather, who do people believe will have the last word? Which one is expected to play chicken?
And that's the problem. Just how do we create the expectation of a little bit of fiscal dominance?
Worthwhile Canadian Initiative 
“I think the truth maybe lies somewhere between.”
Me too. The world seems much more complex than our theories.
I’d agree with your point about Z, as long as it implies that central bank Z had no government bonds on its balance sheet. Because if it did have a few gov’t bonds, and government Z throws away assets, gov’t Z’s bonds will collapse in value, and central bank Z will take a hit to its capital.
On second thought, I guess that the definition of 100% independence implies no government bonds on the central bank balance sheet, and therefore your point about Z makes sense.
Re: Helicopter money
My point is that the central bank has to have an asset on its books to balance the money it creates, and that asset will be a liability on some other entity’s balance sheet. Usually the government’s, backed by the government’s power to tax.
So I’m not sure a government can throw away its assets.
Me, I have anways favoured blimps over helicopters. Money from government always tends to drop in lumps (so you need to consider who gets it when working out what is likley to happen). The helicopter scatters it with unrealistic equity.
That apart, Nick is talking about a consistent way of creating beneficial self-fulfilling expectations. You cannot do that with strict Ricardian expectations about a government that does not directly act to increase future disposable income. But if he expects the actions of the government to increase the PNV of future income by more than the cost of these actions; surely a citizen called Ricardo will now save less than proportionately because he will expect an increase in future taxable capacity?
In the general case, I think that a necessary condition for fiscal dominance to be desirable is confidence that the government will spend the extra money so as to increase expected future economic output above the future economic output expected without the government action. That is to say,(since expectations tend to be self realising) it depends as much on our state of hope or despair for the economy without the government action as upon what we think government action will achieve.
Ruth: How does it go? “Are you now, or have you ever been, a member of any organisation of accountants?” Because what you are saying sounds like the product of a mind that is so dominated by an accountant’s way of thinking, that you say things that would not make any sense at all, otherwise. You illustrate perfectly the dangers of accounting.
“My point is that the central bank has to have an asset on its books to balance the money it creates…” No it doesn’t, if it gives away money for free.
“…and that asset will be a liability on some other entity’s balance sheet.” Not necessarily. It might be a computer, or bricks and mortar, or gold, or land.
“So I’m not sure a government can throw away its assets.” !!!
What if the government takes all its trucks, computers, desks, etc., and just throws them in the garbage? Or gives them to its cronies?
David: “In the general case, I think that a necessary condition for fiscal dominance to be desirable is confidence that the government will spend the extra money so as to increase expected future economic output above the future economic output expected without the government action.”
There’s some truth in what you say here. A few months ago, I did a post on Ricardian Equivalence making what I think might be a similar point. Fiscal policy will be more powerful if people expect the government spending will be on useful investment that will raise people’s future disposable income. But I think that’s a very different point from the one I’m making here, which is that fiscal policy will be more powerful if people expect a future increase in the money supply rather than higher future taxes.
Ruth: If a Govt gives away money, then I would book it as a negative equity entry on the Govt books, so it’s a negative entry on the liability side. This is deficit spending, it happens all the time, the sum of all past actions is the “national debt” and it equals the amount of NFA equity the Govt sector has paid-in to the non-Govt sector. This is why I sometimes refer to is as “paid-out equity” but that has its own issues.
You always need to have the balance sheets balance, but the balance sheet for a currency issuer naturally should look extremely different than a balance sheet for a currency user. The required negative equity position may be the most unusual difference though.