Mankiw admits that we could have a lottery scheme where some currency gets knocked out of circulation randomly – making the return on currency negative. However, he comes up with an interesting point if we just wanted to lower interest rates through a negative OCR.
If reserves earned a negative return at the margin, banks would have more incentive to lend (which is the motivation for these proposals). But more lending might not be the outcome. Banks could instead discourage deposits by, for example, passing the reserve fee on to depositors. Deposits would then earn a negative return, which would give households an incentive to hold currency rather than bank deposits
If the return goes negative for banks they may just pass it on to households who will “horde currency”.
However, I have to ask – will household’s just hold currency as savings? Maybe they will go out and buy bonds or gold or copper – or any sort of durable asset which could double as savings. As long as the asset holder will spend some of the money – then the final impact of the negative OCR will still be stimulatory (either for prices or output depending on your view of the world).
Its like a hot potato situation – banks don’t want to lose money on reserves so they don’t take deposits and charge a negative interest rate, households don’t want to accept a negative interest rate so they buy assets driving up asset prices, the ex-asset holders have cash which they spend.
Now it is possible that households won’t do this, and they will put all the money under their bed – but I just don’t buy it. Not when there are other savings vehicles that will offer a positive rate of return.
By paying the banks to borrow money from the central bank as well, we will get them to go out and buy bonds and the such – doing the same jazz.
Note: Whether we SHOULD do this is a whole other question – the only purpose of this post is to discuss if we CAN do it. There is a world of difference …