Sunday, November 3, 2013

With nothing particular in mind

Sometimes I wake up in the morning with nothing particular in my mind. I get a cup of coffee (or tea. Lately, more tea than coffee. Tea with honey) and sit down at the computer. Often, I end up at FRED.

Today is one of those days.

Graph #1: What's called "Household Debt" ... Debt of Households and Nonprofit Organizations
Blue: Household debt relative to GDP (relative to GDP, for context) (because everybody looks at household debt, and everybody looks at it relative to GDP).

Red: Percent change from year ago, of same.

The blue line rises to a mid-1960s peak... At that point the red line falls below zero (see the right scale) and remains below zero for a while, so that the blue line trends down for a few years.

In the 1970s there is a resurgence of debt growth, debt-to-GDP growth really, with the uppermost peaks of the red line trending uphill until the mid-1980s; and the blue line trending upward again also.

After the mid-1980s peak, the growth of the red line falls back toward zero (right scale). The blue line continues rising, but at a slower rate, through the 1990s.

A large spike in the red line around the year 2000, then a gradual tapering, appears in the blue line as a massive increase from 2000 to the 2009 recession. Decline thereafter in the blue line, as the red line again goes below zero.


I'm wondering if the uptrend from the latter 1960s to the mid 1980s has anything to do with the inflation of those years. Yes, yes, I know: The debt here is "in context". Inflation affected both the debt number and the GDP number, and it should have no bearing on the direction of trend.

Oh -- but that's not really true. I looked at the erosion of debt before. Inflation today, or inflation on a given day, let's say, affects not only prices and borrowing at that moment, but affects also the cost of all pre-existing debt. (I bought a house back then, and after a few inflationary years my mortgage payment took a much smaller bite out of my weekly paycheck.) In those inflationary years, the cost of accumulated CMDEBT eroded as GDP and incomes were inflating. The ratio, the blue line should have been trending down.

But during most of the Great Inflation, the blue line wasn't trending down. It was trending up. That means debt must have been growing faster than GDP -- growing so much faster that it kept the blue line going up despite the inflation of those years.

Have a look at growth rates:

Graph #2: Percent Change from Year Ago for GDP (blue) and CMDEBT (red)
The red peak of the latter 1960s is lower than the blue peak. Household debt grew more slowly than nominal GDP in those years. This corresponds to the decline on Graph #1 between the mid-1960s and the 1970 recession.

Between the 1970 and 1974 recessions the red and blue peaks are nearly the same. The red is a little higher and maybe a little wider. CMDEBT, represented by the red line, grew a little faster than GDP. And sure enough, on Graph #1 the blue line goes up a little between the 1970 and '74 recessions.

Between the 1974 and 1980 recessions the red peak is clearly higher and wider than the blue. CMDEBT grew significantly faster than GDP in those years -- and again, the rising blue line on Graph #1 confirms this.

So yes: Debt was growing faster than GDP in the 1970s, despite the inflation of those years. And no: the uptrend on Graph #1 from the latter 1960s to the mid 1980s doesn't seem to be because of inflation. It seems to be despite inflation.

The 1970s was the time of the greatest excitement of the Great Inflation. Inflation was driving GDP up... Inflation was eroding the burden of debt... But did we take advantage of that erosion to pay off debt at a faster rate? We evidently did not. Debt grew faster in the early 1970s than in the late 1960s, and faster in the late 1970s than the early 1970s as Graph #2 shows; and, as Graph #1 shows, the ratio of CMDEBT to GDP rose through the 1970s. Not only rose, but accelerated.

Did I make an extra payment every so often on my old mortgage? I did not. Instead, sad to say, I started using credit cards. My debt went up faster than my income.


Graph #3 shows "percent change from year ago" growth of household debt (blue) and the rate of inflation as measured by the GDP Deflator (red):

Graph #3: The Growth Rate of CMDEBT (blue) and GDP Deflator Inflation (red)
Inflation (red) peaks twice in the 1950s, reaches a bottom at the end of the 1950s, and starts to climb rather early in the 1960s. This is where the Great Inflation begins.

During those years, the growth rate of household debt is slowing. It slows until the 1970 recession. It slows for the better part of a decade after inflation starts to rise. It was not the growth of debt -- not household debt anyway -- that turned the embers of inflation into a raging fire.

Do I find this disturbing? Of course I do. I need another look. Graph #4 shows the same measure of debt (blue) as Graph #3, and the rate of inflation as measured this time by the Consumer Price Index (red):

Graph #4: The Growth Rate of CMDEBT (blue) and Consumer Price Index Inflation (red)
Same story. Debt growth slows all through the 1960s; inflation picks up early in the 1960s. Why do I find this story disturbing? Because I blame debt for everything! But I cannot blame the slowing growth of household debt for the accelerating inflation.

It bothers me, yes. It doesn't bother me a whole lot, though, because I know there's gotta be costs somewhere that ignited the inflation. It's just that those costs are not to be found in household debt.

For the record, then, let's have a look at some other debt. I'll start with TCMDO, total credit market debt, and subtract CMDEBT out of it. Subtract out the household debt. Come to think of it, I'll subtract out the Federal debt, too. Everybody knows that the Federal debt is the source of all our troubles -- thinks they know -- so let's eliminate that source of troubles along with household debt, and take a look at the debt that remains.

And I'll show it with the red and blue lines both using the same scale, the left scale, so you can better judge how high the rate of debt growth was in the 1960s and '70s, as compared to the rate of inflation:

Graph #5: The Growth Rate of Debt Other than Household and Federal Debt (blue)
and Consumer Price Index Inflation (red)
I rest my case.


Jazzbumpa said...

Re: Graph 5. Slow down a little.

First off, correlation does not imply causation. Second, it's only neat looking from about '60 to '80. It's sketchy during the 50's; and after '80, the correlation totally breaks down.

in fact, during the 80's and 90's the curve motions are almost exactly inverse.

If it's determinative on the way up, then is also has to be determinative on the way down.


The Arthurian said...

Did you notice, the post was written entirely about the mid-1960s to early 1980s.

Obviously when Reaganomics replaced the Keynesianish economics of the golden age, lots of things got different.

Correlation does not imply causation... COST implies causation.

geerussell said...

I don't have a fleshed out position or a real argument, I just want to pop in, yell "real shocks!" and run back out.

The Arthurian said...

Gee, SRW's baby boom story made one hell of an impression on me. It'll be a long while before I get that story worked into my thinking. But I'm not ever going to say that part of economic policy is to make decisions regarding population control.

The price of oil, in the 1970s, was no more than a contributing consequence.

Now, toilet paper and cabbage patch doll shortages, that was real...

Jazzbumpa said...

My Swedish is not so good, alas.

Yes, you're talking about mid 60's to 80's. But if you think there's a relationship between debt and inflation that exists for a while and then doesn't, you have to have some mechanism to explain why.

Take your debt measure times the prime rate as a proxy for cost, and it continued to go up as inflation went down.

Can Reaganomics explain that?


Jazzbumpa said...

To geerussell's point, there were significant supply shocks in the 70's.


The Arthurian said...

Jazz: "if you think there's a relationship between debt and inflation that exists for a while and then doesn't, you have to have some mechanism to explain why."

The piece of the puzzle that I bring to the table is the cost of finance, which competes with wages and with profits. The cost of finance grows with the size of finance. This is conceptually simple, and always true.

How the cost of finance affects the economy is an endless tale that depends on the size of finance, but also on policies and the changes in policy.

The Arthurian said...

Wow Jazz, your g=o0R graph is bullshit. Just now I tweaked it to make the left-hand border units the same... and the differences you show disappeared!