The Interest Rate and GDP Growth (follow-up)

Yesterday, I looked at Rodger Mitchell's "The low interest rate/GDP growth fallacy", wrapping up my criticism with a look at his FRED graph:

Graph #1: Rodger's graph

Of this graph, Rodger said:

Not only are low interest rates not associated with high economic growth, but to a slight degree the opposite seems to be true. There seems to be a small correlation between high interest rates and high GDP growth.

Rodger is not specific about the correlation he reports. But that's okay; I expect to agree with him on this point. Be patient with me while I get there.


In the previous post I said, "Rodger's graph begins in 1971, which is just about the end of the good years of the post-WWII period. So, Rodger's graph is a study mostly of the bad years."

I said, "I looked at what FRED has on this. Rodger could have taken the graph back to 1954. I'm thinking, if the good years supported his argument he would have shown them. But he didn't show them."

I said, one can clearly see on Rodger's graph that high interest rates *did* inhibit growth, "creating the recessions of 1974 and 1980 and 1982."

In sum, I suspect that from 1954 to 1984, Rodger's graph would contradict Rodger's claims; and that after 1984 the graph behaves in a way that does not contradict him.


All through my earlier post, the running theme was that the economy changed because our reliance on credit increased:

The economy's response to interest rate cuts has changed because our use of credit is vastly greater... [If] monetary policy doesn't work as well as it once did, there are reasons for that... In an economy that uses credit for everything, changing interest costs affect everything.

So without really getting into Rodger's graph, I found that what he said of it might be true only of years after 1984, and seemed not to be true of the earlier years. And I offered an explanation for the change: the increasing reliance on credit.

Now, we shall get into Rodger's graph. I reconstructed it at FRED. But I used inflation-adjusted GDP rather than Rodger's nominal, to get a more accurate picture of growth. And I start with the early years:

Graph #2: 1954-1984

The vertical gray bars indicate recessions.

The four right-most gray bars indicate the recessions of 1970, 1974, 1980, and 1982. It is quite easy to see the blue line (the interest rate) increasing on the approach to all four of these recessions. At the same time, in each case, as the interest rate increases, GDP growth (the red line) falls.

Despite what Rodger says, increasing (or, "high") interest rates clearly had a negative effect on growth in the 1969-1984 period.

But with Graph #2 I couldn't get a good look at the early years, before 1969. So I cropped the graph, cutting off the years after 1974, to get an enlarged view of the earlier period. In addition, I gave each trend-line its own vertical axis, to let each one better fit the whitespace of the graph:

Graph #3: 1954-1974

The interest rate (the blue line) increases from 1954 to 1958. At first, GDP growth (the red line) is not restrained. But after it rises above 7.5% (left-hand axis) GDP growth falls, and then falls again with the onset of recession. Rising interest rates caused, or correlate with, a slowdown of growth.

During the recession, interest rates are allowed to fall until growth resumes.

After the 1957 recession, the pattern repeats: Interest rates rise from 1958 to 1960. During that period, GDP growth increases, peaks, and finally falls. The fall of growth continues into the 1961 recession. Again, rising interest rates cause or correlate with a slowdown of growth.

During the recession, interest rates again ease until growth again resumes.

After the 1960 recession, the pattern repeats again. From a low of 2.0 (right-hand axis) interest rates gradually rise to over 8 percent by 1970. During that decade growth rises, then varies at a high level. But interest rates continue to rise, and by 1970 the economy falls into recession.

Again during the 1970 recession, interest rates fall until the recession ends and growth resumes.

A detail from that decade: Around 1966 interest rates rose sharply for a year. At the same time, we saw a sharp drop in growth. (This is the "near-recession" that I sometimes mention.) But before 1968 interest rates fell, and then held steady, and the economy did not fall into recession.

Rowland Evans and Robert D. Novak described this detail in the July 1971 issue of Atlantic Monthly:

Having choked off the money supply as an anti-inflation device in 1966 so tightly that it produced a serious slump in housing and construction (called by some a "mini-recession"), the central bank started pouring out money too quickly and too generously in 1967 and thereby spoon-fed a new inflation.

By 1968, interest rates were rising again, slowing GDP growth and leading to the recession of 1970.

We looked at the years after 1970 in the discussion of Graph #2 above. But Graph #3 makes it evident that after the 1970 recession interest rates stayed low for a couple years while GDP growth again rose rapidly. Then interest rates started rising, GDP growth peaked, and the economy again fell into recession.

The pattern is consistent: Low interest rates allow growth to increase. Rising interest rates eventually cause a slowdown of growth. Beginning in 1954 and continuing for 30 years, Graphs #2 and #3 show that low or reduced interest rates encourage growth, while high or rising rates hinder growth.

Rodger says, "there is no historical correlation between interest rates and economic growth." Rodger is wrong. The correlation *is* historical.


We look next at the years since 1984:

Graph #4: 1984-2011

Rodger says there "seems to be a small correlation between high interest rates and high GDP growth."

"How can this be?" he asks.

"When interest rates are high," Rodger says, "the federal government pays more interest on T-securities, which pumps more money into the economy. This additional money stimulates the economy." In other words, we have come to rely on interest income, as opposed to income from wages and profits.

As for myself, I think his explanation is probably true but not significant. It plays a minor role. There are other factors of greater weight. But I do find it interesting that Rodger's explanation depends upon the concept of a growing reliance on credit.


Rodger's analysis begins with a bad argument about inflation and recession being opposites. He reinforces his analysis by relying on the absurd notion that interest is never compounded, that it is always withdrawn and spent into circulation. And he rounds out his argument by asserting the thing he is trying to prove -- that "there is no historical correlation between interest rates and economic growth."

Finally, Rodger presents a graph of GDP growth and claims that it supports what he has been saying. But the graph does not even show GDP growth. It shows percent change in inflating GDP. Growth cannot be measured while inflation skews the numbers.

My argument is that the economy changed, and that the change is visible in the two different behaviors of the graphs we have reviewed: before 1984, and after 1984.

Rodger's argument is that inflation and recession are not opposites; and therefore monetary policy cannot work. He neglects the fact that it worked like clockwork for the 30 years from 1954 to 1984.

Let it be enough, for now, that Rodger and I agree on this: Since some time around 1984, monetary policy has not worked as well as once it did.

In fact I should prefer to say that raising and lowering interest rates DID make sense in the 1950s and 1960s. It does not make sense any more, or at least it does not *work* any more, because the economy has changed since those days: We use far more credit now than we did then, per dollar of spending. And because our use of credit is vastly greater, the influence of changing interest rates is vastly greater.