Mankiw's basic accounting error

We live in a fiat world, where assets and liabilities have to balance, but Harvard econ profs seem to think we are still on some gold standard, where money is an asset to its holder, and a liability to no one. Greg Mankiw's misconception is clear:

If we want to prop up aggregate demand to promote full employment, what is the alternative to monetary policy aimed at producing negative real interest rates? Fiscal policy. Essentially, the private sector is saying it wants to save. Fiscal policy can say, "No you don't. If you try to save, we will dissave on your behalf via budget deficits." That fiscal dissaving would push equilibrium interest rates upward. But is that policy really welfare-improving compared to allowing interest rates to fall into the negative region?

Private sector savings is funded by Federal deficits. If the private sector wants to save, it will either shrink aggregate demand and fail, or it will be generate larger deficits and succeed. Greg's fiscal policy makes no sense. For the private sector to save the public sector *must* dissave, the same way every asset must be balanced by a liability. It's an iron law of accounting. Public deficits, far from thwarting a private demand to save, are necessary and sufficient to enable that saving.

As for whether fiscal dissaving pushes equilibrium interest rates upwards, surely the example of Japan shows that interest rates respond to whether fiscal deficits are large enough fund private savings demand. Japan's been at ZIRP for years, and public debt runs at 250% of GDP.