Sunday, February 12, 2012

Dancing with Steve

It's getting difficult for me to avoid reading Steve Keen.

In Economics in the Age of Deleveraging Keen identifies the "Age of Leverage" -- for the US, the years between World War Two and 2007 -- and the "Age of Deleveraging" (the years since 2007). He provides two graphs showing dramatic increase in private debt followed by sudden decline, and says "only the Great Depression compares".

And Keen says this:

Non-economists might expect professional economists to pay great heed to these indicators—after all, surely private debt affects the economy? However, the dominant approach to economics—known as “Neoclassical Economics” —ignores them completely, on the a priori grounds that the aggregate level of private debt doesn’t matter: only its distribution can have macroeconomic impacts.

Keen objects to the notion that "the aggregate level of private debt doesn’t matter". In other words, Keen thinks the aggregate level of private debt DOES matter.

He doesn't put it in those words, exactly, but that's what he is saying: The level of debt matters. I like to say the level of TOTAL debt matters; Keen here says it's the level of PRIVATE debt that matters. But anyway total debt is composed mostly of private debt, so we are not very far apart.

So then I was thinking maybe Keen had changed his view since I read him some time back. Actually, since my first post that references Keen. In that post I wrote:

Keen says the "superficially good economic performance" since the mid- to late-80s was "driven by a debt-financed speculative bubble." I say it was driven by the growth of debt. I say the host of Debtwatch has overstated the case.

I was saying it is the level of debt that matters: An excessively high level of debt is the problem. Keen was arguing that the problem was not the level of debt, but the use to which debt was put: speculative use. Keen had explained:

... the debt added to the economy’s servicing costs without increasing its capacity to finance those costs. At some stage, the growth of unproductive debt had to falter, and when it did a serious financial crisis would ensue...

But I thought otherwise:

Keen's explanation is wrong. It suggests that if all the debt was productive debt, there would have been no problem. That's blindly optimistic. Accumulating debt would have created the crisis eventually, no matter how you label that debt.

So. In the older post (June 13th, 2010) Keen said the problem was not debt, but "unproductive debt" specifically. In his recent post, however, he seems to suggest it is the aggregate level of debt that matters. He seems to say that the problem results from how much debt there is, rather than what kind of debt it is. I thought he was coming around to my way of thinking. I was bejabbered.

Bejabbered, but wrong. In the recent post Keen says

an increase in debt adds to aggregate demand—and it is the primary means by which both investment and speculation are funded.

Speculation, again. The level of debt matters to Steve Keen because it influences "both investment and speculation". If not for speculation, he says, things would be grand!

Keen remains concerned about the difference between productive and speculative debt. His remarks regarding the level of debt arise from his concern with the level of speculative debt. Keen still fails to see the cost of debt in general as a cost issue for the economy. A factor cost issue.

Keen, from the recent post:

If the change in debt is roughly equivalent to the growth in income ... then nothing is amiss: the increase in debt mainly finances investment, investment causes incomes to grow, and the economy moves forward in a virtuous feedback cycle. But when debt rises faster than income, and finances not just investment but also speculation on asset prices, the virtuous cycle gives way to a vicious positive feedback process: asset prices rise when debt rises faster than income, and this encourages more borrowing still.

Keen sees no problem with accumulating debt, as long as that debt was created for productive use and not for speculation.

What I see is that debt ALWAYS "rises faster than income" -- except at the end:

Graph #1: Total Debt relative to GDP (GDP is Income!)

If there is any question in your mind that excessive debt can create problems in the economy, check out the Debt Accumulation item in my sidebar....


Graph #2: Total Debt relative to GDP, through 1982

Graph #3: Total Debt relative to GDP, 1962-1972
Well look at that! Sure enough, the ratio went DOWN from 1965 to 1969.

Oh no-no wait! It went up from 1966 to 1968.

So what do we have?  1965-66 and 1968-69 the ratio went down.

THOSE were the years when debt was "productive"???


jim said...

Hi Art,

you wrote:

Keen's explanation is wrong. It suggests that if all the debt was productive debt, there would have been no problem. That's blindly optimistic. Accumulating debt would have created the crisis eventually, no matter how you label that debt.


On its face this comment makes no logical sense.

Productive debt means, by definition, that income (production) grows faster than debt.

If you have people borrowing where the economy always grows faster than the debt how can the debt create a crisis eventually?

Keen's point is that speculative debt by definition will cause real growth to be lower (slower) than the debt. And that eventually means the debt grows to the point that the economy cannot support it.

If you see that debt always rises faster than income, that is because you are only looking at unproductive debt.

The evidence indicates that in the US speculative debt crowded out most productive debt decades ago.


The Arthurian said...


So what are you defending? When was debt ever "productive"? How many decades back do we have to look?

In my view the increasing reliance on credit increases the factor cost of money which makes finance more profitable, and makes labor and non-financial business relatively less profitable (which drives more money into finance)...

CREDIT USE is ALWAYS productive TO THE BORROWER. But this is a micro-economic evaluation that has no bearing on macro-economic conditions.

Debt is NEVER productive. It is always the burden that remains after the ejaculation of credit use is finished.

Anyway, who watches the watchers? Who decides what is "speculative"? Moritz
suggests that borrowing to buy a home is a non-productive use of debt. Maybe we should all live in tents?

jim said...

There is always some productive and some unproductive debt. Right now it looks to be mostly productive. The evidence of that is that GDP is growing faster than debt. Nobody is signing a mortgage today because they believe the price will shoot up. Today borrowers are now making the same calculation they used to make in the 50's that the monthly cost of ownership will be less than their monthly cost of renting. The calculation has no inclusion of any expectation that the asset will increase in price and may even include some expectation of a fall in price.

It seems obvious that
1940-1980 did not have an over-abundance of speculative debt.

My take on that is that there was still a memory of deflation of the 30's and therefore very much less of the willingness to gamble on rising prices. The point Keen makes is that it is the gambling on rising prices that creates these rising prices.

I agree that you can state without reservation that at some point the volume of credit can be bad in the same sense that at some point eating can be excessive.

Jazzbumpa said...

Art -

I come down squarely on Keen's side, and after this post, I'm quite surprised you don't, as well.

Jim as it quite right, and your statement, "Debt is NEVER productive," is quite divorced from reality.

What I see is that debt ALWAYS "rises faster than income"

Have another look at Fig. 1 of the current post; no growth from 1960 through the mid 70's. Very little growth over the first three decades shown.

And where is the break point? right after 1980. Hmmmmm.

Debt that is used for investment to create wealth is clearly good. Keen is right.

Debt that is used for unmitigated rent-seeking is clearly bad. I again cite the unregulated market in derivatives - which contributes 0.000 to real investment - notionally valued at 10 to 25 x aggregate global GDP.

I'm ambiguous about consumer debt. It's a burden on the borrower, but enables demand, which enables businesses, and the economy in general. Here is where ability to service personal debt comes most obviously into play.

And once again, I'll emphasize that debt is a symptom, not the core problem. The problems of rent-seeking, speculation, and low wages leading to debt are the direct results of neoliberal policies.


The Arthurian said...

jfc i'm not arguing in FAVOR of speculation.

Jazzbumpa said...

i'm not arguing in FAVOR of speculation.

I realize that. But you seem to be arguing AGAINST any and all debt, all the time.

That is my point of disagreement.


The Arthurian said...

I'm arguing against excessiveness, actually.

I have a good argument too, I think, about why as the level of debt increases it tends to favor speculation and non-productive uses more and more. But maybe I have not made that argument yet. I can't keep track :)

Jazzbumpa said...

I'm arguing against excessiveness, actually.

I can agree with that. It's tough to draw that line, though.

I really think you should give some more consideration to the causes of indebtedness and the various uses of productive and non-productive debt.


Jazzbumpa said...

RE: the update, which I just discovered.

First off, you are dealing in decimal dust. But, even at that level, you find the YoY change in GDP moving pretty much in lock-step with Debt/GDP.

THOSE were the years when debt was "productive"???



The Arthurian said...

JzB:"First off, you are dealing in decimal dust."

My point exactly. To find a time when income rises faster than debt, you have to chop up the timeline into very small pieces.