Scott Sumner  

What does it mean to suggest that interest rates are "artificially low"?

PRINT
Google and the EU foreign poli... Where Are the Pro-Life Utilita...

I recently received this request in a comment section:

Slightly off-topic, but Scott you would *love* an idea which is becoming more widespread here in the UK, that the housing shortage here is being caused by "artificially low interest rates".

It is being peddled by worryingly-influential right-of-centre finance commentators in newspapers.


Let's start with the term 'artificially'---what does that actually mean? It might mean a shortage or surplus exists, as when New York rent controls artificially hold rents below equilibrium. It might mean monopoly power, as when OPEC artificially raises equilibrium oil prices though output restrictions. This latter use of 'artificially' means something more like monopoly power than price controls. Clearly the role of monetary policy in the credit markets is more like OPEC than NYC rent controls. The fed funds rate does find an equilibrium, it's just that central banks are such big players that they can affect that equilibrium rate.

OK, but how would we establish whether the equilibrium interest rate is "artificially" low or high? There's really only one set of criteria, which (AFAIK) were first modeled by the economist Knut Wicksell. He referred to the natural interest rate as the rate that would lead to macroeconomic equilibrium. He used price stability as the benchmark for equilibrium, although in modern central banking the analogous criterion would be something like 2% inflation. Because Britain and other major developed countries are currently experiencing less than 2% inflation, there is no evidence that rates are artificially low. If they were, inflation would be much higher.

Some people argue that supply-side factors such as competition from China are holding down inflation, and that the inflation shows up in asset markets like housing and stocks, instead of consumer prices. That's a good argument in principle, but is it true? George Selgin and others have persuasively argued that something like NGDP growth is a better criterion for monetary stability. Selgin suggests that rapid productivity growth should lead to lower inflation, and that if central banks artificially maintain something like 2% inflation in an environment of fast productivity growth, then the economy might become overheated and distortions would occur. I agree.

Unfortunately most proponents of the "artificially low interest rates" hypothesis don't seem to have a coherent macro model underlying their arguments. If China competition were actually holding down inflation, we should see unusually fast productivity growth. But in the UK there has been almost no productivity growth in the past decade, and hence pointing to supply-side factors actually deepens the puzzle. If rates were artificially low in the UK, then we ought to see either high inflation or high productivity growth. But both of these variables are growing very slowly.

There is absolutely no mechanism by which monetary stimulus could boost asset prices without boosting NGDP. None.

If rates were artificially low, then attempts to raise them should be successful. But recent attempts by central banks to raise rates have all gone poorly. Japan tried in 2000 and 2006, Sweden in 2010, the ECB in 2011. In each case the economy soon went back into recession and/or deflation, and the central bank had to cut rates back to zero to prevent an outright depression.

When I studied macroeconomics at the University of Chicago in the late 1970s, I felt we were ahead of the rest of the profession. One advantage was that we never reasoned from an interest rate change. Friedman emphasized that low rates did not mean easy money, and that you needed to look at other indicators (he preferred M2, and then later in his life inflation.) It's very discouraging to see things flop-flop in 2015, where you now tend to see people on the right drawing all sorts of unjustifiable conclusions from the low level of interest rates. In contrast, many economists on the left recognize that low rates reflect a weak economy and low inflation, as well as possible long-term "stagnation" factors. I guess we'll have to keep fighting this battle over and over again.

PS. I have some comments on interest rates at a new EconTalk interview with Russ Roberts. A commenter pointed out that I didn't clearly enough distinguish between ex ante and ex post returns on capital. I should have emphasized that it is ex ante returns that matter.


Comments and Sharing


CATEGORIES: Finance , Monetary Policy




COMMENTS (13 to date)
ThomasH writes:

Of course we know what "artificially low" means when used by those who use it -- rates that are too low to choke off the meager recover we are experiencing.

What might an intelligent use mean? I'd say it would be a rate, influenced by monetary policy, which is lower/higher than what it would if monetary policy were optimal. For example, during 2008-15 monetary policy had not yet returned ngdp to its trend growth rate so long term borrowing rates that government should use for determining investment allocations were "artificially high."

Of course different criteria for judging the optimality of monetary policy (which is also partially dependent on the way fiscal policy reacts to monetary policy) would give different judgments about whether any given rate were artificially high or low.

Jan writes:

Regarding interest rate too low, how about the spread between mortgage rate and deposit rate being too high? How would that be detected, and has the presence of Fannie Mae and programs specifically designed to give credit to worthy individuals like like veterans without down payments affected this spread?

If the spread is bigger than what is justified based on risk of default, risk of early down payment and risk of future inflation, is this a good opportunity for individuals too pool their money, and issue loans to the highest bidder in terms of interest rate?

I am not an economist, so my apologies if my terminology is inconsistent, the proposed arrangement above seems like what is described as shadow banking, and I am somewhat surprised it is not more prevalent. I suppose shadow banking is uninteresting to macro economists, as money supply is unaffected, but in terms of not feeding middle men who benefit from seigniorage, this seems like an good analog to the recently outlawed practice of healthy individuals coming together for health insurance, and excluding those expected to increase costs, like those preferred groups who get by without down payment.

Having an 800+ credit, I don't think I get the low spread my near zero risk of default should warrant.

Devin Helton writes:
Let's start with the term 'artificially'---what does that actually mean? OK, but how would we establish whether the equilibrium interest rate is "artificially" low or high? There's really only one set of criteria, which (AFAIK) were first modeled by the economist Knut Wicksell. He referred to the natural interest rate as the rate that would lead to macroeconomic equilibrium.

The natural price of anything is the price based on the supply and demand of a free market. The interest rate is simply the price of future money. The natural interest rate is the rate that equalizes supply and demand between those who wish to consume or spend on capital goods now and those who are willing to forgo the current use of money for greater consumption later.

When people like me argue that interest rates are "artificially low," we are not to saying anything particularly novel or interesting. Interest rates are massively influenced by both the actions of the Fed and Byzantine set of subsidies from the rest of the government, (such as the state sponsored agencies that sop up mortgages and student loans, government mandates that require pension funds to own bonds, etc.). Thus all interest rates in the modern financial system are "artificial." Particularly low interest rates are thus "artificially low."

If rates were artificially low in the UK, then we ought to see either high inflation or high productivity growth. But both of these variables are growing very slowly.

There is no fundamental reason that this needs to be true. The existing financial system is massively complex. So to understand the mechanics, let us use a simplified model:

Imagine an economy where the government prints money to increase the money supply by 2% a each year. The government issues the newly minted money in the form of tax rebate checks. There is no Fed nor any other government intervention in lending. This is a very simple free market economy. Interest rates sit at 5% due to the intersection of supply and demand for future money.

Now the government changes policy. It still prints out 2% in new currency each year. But it now distributes the new currency by lending it out, by auctioning off loans. If the government injects massive amounts of money on the supply side of a market, it will drive the price down. By injecting massive amounts of cash into supplying new loans, the government drives down the price of loans, drives down the price of present money. Interest rates will fall, interest rates will be "artificially low." Yet, the same total quantity of new money is injected into the economy. In this model, the new money being loaned out comes at the expense of money that would have been refunded to tax payers. After this intervention, average tax payers no longer get nice tax refunds with which to remodel their kitchen. Instead, they will borrow money at these super low rates to remodel their kitchen. So the same amount of money chases the same total production. Neither productivity nor inflation rises. All in all, interest rates are much lower, interest rates are artificially low, but the overall measures of the economy remain the same.

If rates were artificially low, then attempts to raise them should be successful. But recent attempts by central banks to raise rates have all gone poorly.

The Fed cannot raise nominal interest rates, because the rates have been too low, for too long. It is as simple as that -- existing firms and people are addicted to the current rates. Raising the rates will be like quitting heroin.

This has gotten too long, so I will continue the explanation on my own blog, if you are interested in explanation of how exactly the low interest rates have created dependency: https://intellectual-detox.com/2015-04-28/why-low-interest-rates

Melbot87 writes:

"If rates were artificially low in the UK, then we ought to see either high inflation or high productivity growth. But both of these variables are growing very slowly."

Wouldn't the argument be - not that inflation or productivity is necessarily high (in a general sense) - but higher than it otherwise would be absent the lower interest rates?

So for instance, in the US, if the mortgage market is falling but the US government buys mortgage backed securities to prop up the market and lowers rates to encourage more borrowing, then housing prices could still be falling but just not as drastically as they would have been absent the intervention.

So I don't necessarily see why NGDP would have to rising to show it has an effect.

Brendan riske writes:

Scott, you butchered this. Too blinded to see that this is federal reserve policy. The easy money policy is driving this. The inflation that you expect to see is not being captured in our current calculations. The feds easy money has vastly inflated finacial assets. Now regular people are feeling it's effects in Healthcare and housing. Go out into the real economy and see for yourself.
And you never really explained negative rates. Interest rates cannot possibly be negative naturally! You said it yourself, the first place to see negative rates was japan. Why? It was the first place to ease!
As for your attack on piketty, your dead wrong returns to financial assets, which are owned by the wealthy primarily, are super high. Why? Easy money driving their price up through carry trades! Look at what easy money does to fiancial markets and tell me that it isnt driving asset values up. Savings rates, which average people get, are incredibly low. That leads to income inequality

Jose Romeu Robazzi writes:

Prof. Sumner,
Could you please elaborate on this claim:
"There is absolutely no mechanism by which monetary stimulus could boost asset prices without boosting NGDP. None." ?

It could be well that all the money printing went to buy financial assets, sold by other investors that just wanted more cash ...

Scott Sumner writes:

Traveling today so just a few quick comments.

Government programs that create moral hazard (FDIC, GSEs, etc.) can distort interest rate spreads, and boost asset prices.

Don't equate low rates with easy money. A policy of easy money leads to very high interest rates, often double digit rates. We saw that in the late 1970s and early 1980s.

If the government is understating inflation then they must be overstating RGDP growth, in which case the argument is the same.

Brendan riske writes:

A few things,

I think you were correct when you said that there is a demand issue. I agree that there are limited opportunities to generate returns in the real economy. Is it possible the rgdp growth has been negative? That might explain the lack of inflation.

I think the channels of monetary policy are key to understand what is going on. Massive excess reserves have been injected into the banking system. It has not been lending to real business, because the returns are too low. Instead that money poured into the financial sector. Bond prices are up almost universally, with interest rate on them correspondingly low. The stock market indices are going up far in excess of growth potential or earnings. Also, this money has fed bad merger deals. speculation and carry trades are the order of the day. All of this largely feeds the wealthy, who own the assets appreciating. Since they do not spend money in the same way, and don't proportionally consume in the same way, it hasn't shown up majorly in cpi. But inflation is there in the assets. Also don't forget that much of the expansion of the usd supply was absorbed overseas. As dollars pour in from abroad, i think it will increase inflation
There are many bizarre examples right now which I don't see possible without acknowledging that government policies, especially the fed but also in other countries,have depressed interest rates by giving money at 0% to the banks. Why else would you lend to the near insolvent European countries at a negative rate?

Luciom writes:

I agree with most of your article especially the point of people arguing about "art. low IR" without a coherent macro model.

But i think you can't dismiss the asset price inflation part of the discussion with a single phrase.

Quoting:

There is absolutely no mechanism by which monetary stimulus could boost asset prices without boosting NGDP. None.

I'm not sure this is true, and even if it is you need to elaborate much further.

Unless you believe that monetary stimulus, even when prolonged for a lot of years, has ZERO effect on the risk-free discounting rate in assets' pricing models, then no, you can't say that there is no mechanism by which mon. stimulus could boost asset prices withouth boosting NGDP.

Because if somehow mon. stimulus can lower the 5y or 10y rate (like when you buy half of the oustanding 10y+ t-bonds) then yes, asset prices will move accordingly even if NGDP doesn't change.

With respect to housing prices and inflation: The most widely EU inflation measure (the harmonized index) does not include house prices at all. It excludes them for the most of European reasons: different countries had different standards, so when they harmonized them and couldn't get an agreement on how to fold them in, so they were left out.

In some countries at least, I believe it has been excluding *deflation*, though, as house prices have gone down.

Spain house prices: http://www.tradingeconomics.com/spain/housing-index

EU level: http://ec.europa.eu/eurostat/statistics-explained/index.php/Housing_price_statistics_-_house_price_index

In the US, off the top of my head, I think housing (mostly imputed rents) are about 1 third of the basket used for the CPI. So, the effect of excluding housing can be very large.

*

How's this for candidate for "most Sumnerist statement of the year":

In the euro area, artificially low interest rates are associated with a fall in house prices.

(because the root cause of both is the "artificially" hard money policies of the ECB & don't reason from a price change).

GB writes:

So trillions and trillions of bond purchases across the globe are not evidence of a expansive monetary policy...because interest rates (on those bonds) are low? Sure it makes sense?

danan writes:

[Comment removed pending confirmation of email address. Email the webmaster@econlib.org to request restoring this comment. A valid email address is required to post comments on EconLog and EconTalk.--Econlib Ed.]

Lorenzo from Oz writes:

GB: Is that a flight to safe assets? If so, what does that suggest about monetary policy. (Think aggregate demand and income expectations.)

Comments for this entry have been closed
Return to top