Tuesday, 29 March 2011

Abba Lerner’s flawed ideas on interest rates.

Note: this post replaces a post entitled “Abba Lerner’s interest rate mistake” (15th Dec 2010) because I made a mistake in the latter post (details below). I’ll delete the latter post in due course.


The first two sentences of a paper by Stephanie Bell say “In 1943, Abba Lerner wrote an essay entitled Functional Finance and the Federal Debt. The essay elucidates, in just fourteen pages, the principles that Lerner believed should guide the government's budgetary policies.”

I basically agree with Lerner’s paper, but not with a point he makes about interest rates. By the way, while his essay consists of only fourteen pages, I’ll use the page numbers from the journal in which the essay appears (38 – 51). Also, I’ll use the word “government” in reference to “government and central bank combined”.

There is a paragraph at the bottom of Lerner’s page 40 which starts “The second law of Functional Finance is that the government should borrow money only if it is desirable that the public should have less money and more government bonds…This might be desirable if otherwise the rate of interest would be reduced too low . . . and thus induce too much investment, thus bringing about inflation.”

(In the above mentioned 15th Dec post I thought Lerner was advocating interest rate adjustments so as to optimise the amount of investment: in fact, as the above quote makes clear, he advocated such adjustments so as to control inflation. Doh!)

Anyway, why did Lerner think that adjusting interest rates is desirable as a means of controlling inflation when he has just advocated adjusting government net spending as a method of controlling demand and inflation? (See bottom of his p. 39)

Adjusting interest rates almost certainly influences inflation: e.g. if an economy is near capacity and inflation looms, then raising interest rates will presumably result in a finite reduction in investment spending which in turn will reduce inflation. But the important question, which Lerner does not address, is why do we need two tools to do one job? Having two tools doing one job is not really consistent with the Tinbergen principle.

If the two tools have obvious and different merits and demerits, then there could be a case for using both. For an example of a demerit, interest rate adjustments are probably distortionary in that they work only via entities that are significantly reliant on borrowed money. Plus such adjustments exacerbate fluctuations in demand for capital goods, which is not a clever move, because it is precisely instability in the demand for capital goods (via the so called “accelerator”) which is one of the main causes of instability for the economy as a whole.

Fiscal policy can act quickly and needn’t be distortionary.

In contrast, changes in government net spending needn’t be distortionary: e.g. a payroll tax change influences the spending habits of ALL employees, and that is a big proportion of the population. Hence the effect should be fairly non-distortionary.

The fact that in practice changes in government net spending often ARE distortionary and/or take time to implement is irrelevant.

Distortion often arises because for example politicians lobby for their favourite and bizarre bits of pork every time there is money available with which to purchase pork. But it would be perfectly feasible to have SOME fiscal changes (e.g. payroll tax adjustments) which are relatively free of distortionary effects.

Moreover, such distortionary effects as ARE inherent in a payroll tax change can be rectified. For example, as just mentioned, a payroll tax change affects only those in work, which leaves out for example pensioners. But this defect is easily enough rectified in countries with a state pension scheme by temporarily changing the state pension. Indeed, this sort of measure is already operative in the U.K. in that pensioners get a variable Winter fuel allowance, depending on how cold the Winter is.


A similar point applies to the SPEED with which fiscal changes can be made. It is perfectly reasonable to spend months if not years debating some fiscal changes before they are implemented. The issues involved can be complicated. But that is irrelevant. The important point is that, again, it would be perfectly feasible to have SOME fiscal changes that can be altered quickly when the need arises. Indeed, the U.K. changed its sales tax (VAT) twice during the recent credit crunch.

Possible merits in interest rate adjustments.

As to a possible merit in interest rate adjustments, it might be that interest rate adjustments work QUICKER than fiscal adjustments, in which case there would be a case for using both tools. To illustrate, given a sudden need to boost demand, interest rates could be dropped so as to obtain a QUICK change to aggregate demand, while changes to government net spending (fiscal changes) could be implemented at the same time and would eventually take over the “work” done by interest rate changes.

But the evidence seems to be that interest rate changes do not work all that quickly. At least according to the Bank of England, such changes take about a year to work. See 8th para here.

As to how quickly fiscal changes work, the evidence seems to be “quickly”. At least where a fiscal change results in a change in employees’ take home pay, one would expect some sort of effect almost immediately: that is if the average employees’ take home pay rises by $X in say mid June, I would expect at least SOME of that extra to be spent before the end of June.

And the evidence supports this. The following studies indicate that roughly a half of tax rebates or other windfalls are spent by households within about six months. See here, here, here, and here.

Other problems with interest rate adjustments.

Apart from the fact that interest rate adjustments do not work quickly, such adjustments have other problems, two of which are as follows.

First, in the absence of government interference with interest rates, there is presumably some free market rate of interest. And this rate, in the absence of market imperfections, will maximise GDP. Thus there are costs associated with such interference.

Second, central bank instigated changes in interest rates seem to have no effect on rates charged by credit card operators.


Abba Lerner was wrong to advocate TWO tools for controlling demand and inflation: changes to government net spending AND interest rate changes. That is, the main tool which he advocated, changing government net spending ALONE ought to do the job, though perhaps there’d be nothing wrong with having interest rate changes as a back-up, and second best tool, for use only in emergencies.

Afterthought – 6th April, 2011. On second thoughts, the claim that Lerner thought government should adjust interest rates so as to bring the optimum amount of investment does not seem to be so stupid after all: David Colander in “Functional Finance, New Classical Economics, and the Great Great Grandsons” also seems to have thought that this was what Lerner though, though Colander does not cite any sources to back up his claim. See his p.2, item No.2 under the heading “The Rules of Functional Finance”.

Anyway, the big flaw in the idea that governments should fiddle with interest rates so as to bring the “optimum” amount of investment is that it assumes politicians and bureaucrats actually know what this optimum is. The idea is laughable. Moreover, even if interest rates are a percentage point or two above or below this mysterious “optimum”, it simply means we would get very slightly less or more than the optimum amount of investment. The effect on living standards and GDP would be minimal.

No comments:

Post a Comment

Post a comment.