There’s currently a debate about whether or not the Fed should adopt a policy of “negative interest rates”. But what does that mean? Market interest rates are not a policy; rather they are influenced by various monetary policy choices, perhaps over a long period of time.
Here it will be useful to distinguish between negative market interest rates and negative interest rates on bank deposits at the central bank. Most people ignore that distinction, because (AFAIK) at the moment all the countries that have negative market interest rates also have negative interest rates on excess bank reserves (negative IOER). And yet only negative rates on bank deposits at the Fed can be viewed as an actual “policy”, and even that policy tool is only one component of a broader policy mix. Negative IOER is essentially a tax on excess bank reserves, which are currently assumed to have negative “externalities”.
Negative IOER need not coincide with negative market interest rates. Consider back in 2006, when Treasury bills yielded about 5% interest and the Fed paid a zero percent interest rate on bank reserves. That’s quite a large gap. Thus the interest rate on bank reserves is not necessarily anywhere near to the short-term market interest rate.
Now suppose that in 2006 the Fed had decided to cut the interest rate on excess bank reserves from 0% to negative 0.25%. What would have happened? Actually, almost nothing would have happened to market interest rates. At the time, excess bank reserves were tiny, less than two billion dollars. And 0.25% of that amount is less than $5 million/year, for the entire banking industry. Pocket change. Market interest rates on T-bills would have stayed at around 5%, unless the Fed simultaneously did some other change in policy.
Today, if the Fed were to decide to pay a negative interest rate on excess reserves (and I don’t have strong views either way), then they should do so with the intent of raising long-term bond yields. If the negative IOER policy did not succeed in raising long-term bond yields, then the rest of the policy mix (forward guidance, QE, etc.) would likely have been inadequate. There’s no point of doing negative IOER unless you create expectations of faster NGDP growth, which will generally raise long-term interest rates.
Countries that have negative IOER almost always have excessively tight money, which leads to expectations of very slow growth in NGDP, and this leads to very low or negative long-term bond yields.
Negative IOER is an expansionary policy, considered in isolation. In contrast, negative market interest rates are usually a sign that money has been too tight in the past.