Working toward a definition

A recession is when your neighbor loses his job.
A depression is when you lose your job.

- Ronald Reagan

Following up on mine of the 20th. In The trade cycle: debt is trade, Nick wrote:

If you look at business cycles this way, as a trade cycle, in which the volume of trade rises in booms and falls in recessions, it is totally unsurprising that the volume of borrowing and lending should also rise in booms and fall in recessions.

So I wanted to look at that.

I want to look at debt other than Federal debt (because Federal debt by design is counter-cyclical). I'll look at "percent change from year ago" values because "borrowing and lending" is the change in debt, and "percent change" shows it clearly.

Graph #1: Percent Change from Year Ago, Debt Other than Federal Debt

The blue line goes up and down, up and down like something from a nursery rhyme. But if you notice, every time the line goes down, it crosses a gray bar of recession. Every time except 1967, when there was a "near-recession" that doesn't qualify for a gray bar. But you could know by the blue line that something happened there in 1967, some sort of slowdown in the economy.

So yes, Nick is right: Borrowing and lending rises in booms, and falls in recessions.

Now, look at the graph again. This time, look at the vertical scale, over on the left. The blue line is almost always between 5 and 15. That is, the smallest increase is about 5%, and the biggest increase is about 15%.

But even a lowly 5% increase is still an increase in debt. So what this graph shows is that debt is always increasing -- sometimes quickly, sometimes slowly, but always increasing -- from the first blue pixel of the line to the  last recession bar, the fat one just before 2010 there on the right.

See that low spot in the blue line? Just at the 2010 mark, the blue line has a sharp bottom that reaches down to the -5 level. That's a -5% increase. But of course a negative increase is really a decrease. In this case, a 5% decrease.

On that whole long blue line, the only place that debt actually got smaller was in that deep V. Starting in the 2009 recession when the blue line went below the 0 level, and lasting until the blue line went up above the 0 level again, the size of accumulated non-Federal debt actually got smaller.

In the 60-plus years shown on the graph, there was a brief hiccup of less than three years when debt actually decreased in size. (And that doesn't include the Federal debt!)


So what Nick tells us is that debt increases rapidly in a growing economy, and debt increases slowly during recessions.

Now here is my definition. It is the definition of economic depression: Debt increases slowly during recessions. Debt decreases during depressions.

Simple. Obvious.


Now let's check my definition against the facts.

Graph #2 is based on Series X398 from the Bicentennial Edition of the Historical Statistics. That reference work provides values for Public and Private debt for the years 1916-1970. I'm figuring Private debt is roughly equivalent to debt other than Federal debt. The graph shows percent change for the annual values provided:

Graph #2: Percent Change in Private Debt, 1917-1970
The Google Drive Spreadsheet is available

It's not perfect, but it's damned good. The "percent change" numbers are negative for the years 1930 through 1935, 1938, and 1945. All are depression years except 1945.

So for the years 1917 through 1970 this graph shows one exception. On the FRED graph there were no exceptions -- assuming we're in a depression since 2008, of course.

Overall, for the years 1917 through 2012 -- more than 90 years -- there is only one exception, one outlier that could let you raise an eyebrow to my definition of economic depression.