The flaws in negative interest rates.

Summary.

While negative interest rates probably raise demand, they are not an efficient or “GDP maximising” way of doing so: it’s better to simply have the state create more base money and spend that into the economy and/or cut taxes, while interest is left at a zero or slightly positive level. Two of the reasons negative rates are not efficient are thus.

First, negative rates in theory make negative output viable: not what we need. At the very least, they make “very low return on capital” forms of activity viable.

Second, governments or “states” in effect offer a service to everyone, namely what amounts to a savings bank for those who want a stock of base money (or, much the same thing, government debt). GDP is maximised where goods and services are offered at a price which is close to the cost of supplying those goods and services. And the cost of offering the latter “savings bank” service is near zero. Ergo to maximise GDP, there should be little or no positive or negative interest paid to / charged to those with accounts at the “savings bank”.


Introduction.

The reason normally given for negative rates is that rates are now so low that central banks do not have much scope for cutting rates come a recession, without going for negative rates. There is actually no need for negative rates because, as pointed out by MMTers, government of a country that issues its own currency is master of all it surveys: it can implement any level of stimulus it likes and combine that with any rate of interest it likes.

The latter “MMT” policy DOES require cooperation between fiscal authorities (i.e. politicians) and the central bank. But it’s hard to see anything wrong with two arms of government cooperating. Plus Bernanke backs the “cooperation” idea in this Bloomberg article. As for the idea that that cooperation might result in politicians getting too near the printing press, politicians can actually be kept away from the press under a “cooperative” system. Reasons are below.


Do negative rates actually work?

One problem with negative rates is that they may not actually be an effective anti-recessionary tool at all. Reason is thus.

The main idea behind negative rates is that if people are charged in proportion to their stock of cash, they’ll try to spend away that stock, which in turn should boost demand. But there’s another possibility: people may have some target amount of cash they want to hold, so if X% of that target is taken away from them, they’ll respond by saving so as to return their stock of cash to their target. And the effect of saving is to REDUCE demand!


Maximising GDP.

But a more fundamental problem with negative rates is that they may not maximise GDP. Reason is thus.

It is widely accepted in economics that GDP is maximised where products are available at a price which is near the cost of production. It makes sense to have the price of gold hundreds of times higher per kg than the price of steel: reason is it costs far more per kg to produce gold. Obviously each firm aims to make a profit, but competitive forces keep those profits down to acceptable levels normally, which results in the price of most products being near the cost of production.

Now what’s the cost to the state (i.e. government and/or central bank) of maintaining a stock of cash for each household, firm, etc? It’s almost nothing!

Incidentally I’m assuming that the state does actually make savings accounts available for all. That’s not quite the reality, but it very nearly is. E.g. in the UK there’s a state run savings bank “National Savings and Investments”. NSI provides some of the services available from a regular bank but certainly not all of those services. Plus most people in most countries are free to buy government debt, which amounts to much the same thing as putting money into a term account at a state run savings bank. (NSI actually invests only in UK government debt).

As Warren Mosler and Martin Wolf (chief economics commentator at the Financial  Times) said, government debt is almost the same thing as base money.

So to keep things simple, let’s just assume the state has a savings bank for anyone wanting to keep a stock of base money. It would clearly be OK for such a bank to charge for ADMINISTRATION COSTS, as indeed existing commercial banks do.

But there is no obvious reason to charge interest on the money in such accounts, i.e. charge negative interest, if the price of the product is to be close to the cost of making the product available. Moreover, and as regards positive rates (i.e. paying interest to those holding state liabilities) Milton Friedman and Warren Mosler argued that the natural rate of interest is zero: that is, they argued that there is no reason for government to issue interest yielding liabilities. So to summarise, if we ignore administration costs, there are good reasons for thinking that neither positive nor negative interst should be paid on state run savings accounts, i.e. on “state liabilities”.


Stimulus.

But assuming the latter “zero interest” policy is adopted, how then do we implement stimulus? Well that’s easy: have the state print money and spend it (and/or cut taxes). That way the private sector ends up with a bigger stock of cash (base money to be exact), which induces the private sector to spend more. And that’s the form of stimulus Friedman advocated in his 1948 paper ”A Monetary and Fiscal Framework…”. Though he advocated the same unvarying amount of stimulus per year, an idea not widely accepted nowadays.

And if we particularly want to raise interest rates at some given level of demand (e.g. the “full employment” level of demand), that is also easily done: just “print and spend” as above but to an excessive extent, and then deal with the resulting excess demand by raising interest rates (which can be done by having government or central bank, i.e. “the state”, borrow more). That of course is not consistent with the Friedman / Mosler claim that government should pay no interest on it’s liabilities. But never mind: the point is that if a government PARTICULARLY WANTS raise interest rates, it has the power to do so. And indeed there could well be an argument for doing that in an emergency: e.g. if there was a serious outbreak of Greenspan’s “irrational exuberance”, then demand might have to be reined in in a hurry, and an interest rate hike would be one way of doing that.


Dispose of interest rate adjustments?

It was implied just above that interest rate adjustments should be abandoned (except perhaps in emergencies) and that stimulus should be done by adjusting the amount of new money created and spent. That certainly conflicts with the conventional wisdom, but it actually makes sense for the following reasons.

First, come a recession, there is no obvious reason why a fall in borrowing and lending is necessarily the cause of the problem: the cause could be a fall in some other element of aggregate demand, e.g. consumer confidence or exports. And second, even if borrowing and lending HAVE FALLEN, that could be for perfectly good reasons.

Indeed, a classic example of that was the bank crisis ten years ago: that was sparked off by excessive and irresponsible lending, followed by a sudden realization by banks that they’d lent to too many no hopers (e.g. NINJA mortgagors). So the sensible thing for them to do was rein in loans. But that caused a recession. So what did central banks do? They cut interest rates so as to encourage more lending: exactly what wasn’t needed! Might as well try curing an alcoholic by giving him crates of free whiskey!


A free market would boost the monetary base in a recession.

Another point in favor of largely abandoning interest rate adjustments and implementing base money adjustments instead is that that is what would happen in a perfectly functioning free market. That is in such a market, the price of goods, services and labor would fall in a recession. That in turn would mean a rise in the real value of the monetary base, which in turn would encourage spending.

That phenomenon does not work too well in the real world because as Keyes put it, “wages are sticky downwards”. But never mind: instead of raising the value of each dollar making up the base, the NUMBER OF dollars making up the base can be increased instead.


Negative output.

Having claimed above that GDP is not maximised where employers make a charge for a product which is not related to costs, it should be possible to point to exactly how that failure to maximise GDP comes about where the state makes an unwarranted charge to those holding state liabilities, i.e. charges negative interest. Well here goes.

Say I can borrow at minus 4%. I could then buy 100 houses, keep them for a year, burn down two of them just for fun, sell the remaining 96, and come away with a 2% profit! Crazy: that amounts to what might be called “negative output”.

Of course that’s a silly example. But there is a serious point there: under a negative interest rate regime, forms of negative output or “wealth destroying” output then become viable. At the very least, forms of output become viable which would not be viable at the Friedman / Mosler zero rate, or a higher rate.


Politicians and the printing press.

Having suggested it  can be an idea for the state to simply print money and spend it and/or cut taxes, there is an obvious problem there namely that that requires cooperation between the central bank and politicians, thus politicians get closer to the printing press. Does that matter?

Well the evidence seems to be that in most countries the degree of independence of the central bank (i.e. how close it is to politicians) does not influence inflation – see chart below which is taken from an article by Bill Mitchell.

However, like most people I suspect, I’d rather keep politicians away from the press. And doing that under a regime where it is possible to “print and spend” is not difficult. All that needs to be done is to give the responsibility for determining the amount to be printed to the central bank (or some independent committee of economists), while politicians retain the right to determine what proportion of GDP goes to public spending. Indeed, that’s not vastly different from the existing system in that an independent central bank has the last word on how much stimulus there shall be: e.g. under the existing system, if the central bank thinks politicians have implemented too much fiscal stimulus, the central bank well negate that by raising interest rates.

Under a system where the central bank determines the amount of stimulus, but stimulus is effected simply by creating and spending new base money (or cutting taxes), the central bank might say “public spending needs to exceed tax by 2% of GDP this year and here’s the money that enables that to be done”. Politicians would then have the choice as to whether to effect that by having public spending equal to 30% of GDP and tax equal to 28% of GDP. Or they could go for 40% and 38%, etc.

And what do you know? That’s exactly the system advocated by Positive Money, the New Economics Foundation and Prof Richard Werner in this work. Bernanke also expressed sympathy with that sort of system. To be exact, in this Fortune article, Bernanke suggested the central bank should determine the AMOUNT of money to be created and spent, while (as suggested above) politicians take the essentially POLITICAL decisions: whether to distribute the largess in the form of more public spending or tax cuts, and if the former, how much goes to education, health, defence, etc. See para starting “A possible arrangement…” halfway down.