Buiter compares the properties and usefulness of government fiat currency, and other intrinsically worthless assets like Gold and Bitcoin.
He reports that only in models with flexible prices can an economy suddenly spit out a zero value of a fiat currency. Noting that in the real world, the prices of goods in terms of government currency are sticky, yet the relative price of goods and gold or Bitcoin are flexible, he offers this as evidence supporting that the value of fiat currency is more assured than that of the others.
Two notes of caution. First, although he doesn’t say precisely what results he’s referring to, I am sure he is talking about models that invoke rational expectations. So in that respect they are not to be taken as comprehensive ways to adjudicate on the relative properties of these different currencies. [Buiter himself notes that sticky prices are sometimes justified on the grounds that agents are less than ‘rational’ in the (confusing) sense typically meant in this literature].
Second, I don’t think Buiter’s observation that the ‘world is Keynesian’ – relative price of government fiat money and goods is sticky – can be taken to be the final word. It may well be, but we don’t really understand why if it is, and don’t know what models that encoded a better understanding would tell us about this. For example, it also seems to be that prices are sticky until they aren’t! Extreme monetary events have greatly increased price flexibility. [witness the unidad de fomento in Chile, with prices indexed, or most hyperinflations or instances of dollarization].
So, if that government-printed note is burning a hole in your pocket, spend it, and don’t miss out on a Black Friday bargain just because a few sticky price monetary models make you think you are safe.