Monday, July 18, 2011

Wrong again!

Time goes by, and mistakes get buried. And sometimes, foundations get built on top of them, and that's no good.

Jim tells me I have an error. I brushed it off the first time. I'm no economist and I do not claim to have any of the skills that economists have, such as the ability to pick the most appropriate numbers to use for any particular graph. For my graphs I'm concerned most with trends and tendencies, not with particular values.

Still, if there is a chance to improve what I do, I ought to jump on it.

This is the graph that Jim objects to:

Graph #1: The Non-Federal Relative, 1949-2010

This is what he offers as the correct version:

Graph #2: Jim's Non-Federal Relative
The maximum Y-axis value on my graph is half the maximum on Jim's. Quite a difference.

Jim's graph uses "Debt Outstanding Domestic Nonfinancial Sectors - Federal Government Sector" (FGSDODNS) from FRED. Mine uses "Gross Federal Debt" (FYGFD).

On mine, the most recent value for Federal debt is 13528.8 for 2010, which is on-track with the $14.3 trillion that people talk about today.

On Jim's, the most recent value for Federal debt is 9569.59 as of the first quarter of 2011. Well below the 14.3 number.

The thing is, I fear Jim is right. I am using numbers from two incompatible compilations in order to use the $14.3 number. What I need, really, is a different version of total debt, one that is compatible with the $14.3T Federal debt number. Gee, I thought total debt meant "total" debt. Silly me.

But for now, this is my Non-Federal Relative and Jim's, on the same graph.

Graph #3: Jim's (blue) & mine (red)

The trends and tendencies are comparable. But Jim's is definitely higher, especially during (and after) the "macroeconomic miracle" years of the late 1990s.

This is all somehow eerily familiar.

For ha-ha's I took Graph #3 and put each line on a separate vertical axis. I thought it might help to "make the two series comparable," to use Milton Friedman's phrase.

Graph #4: Looking for Similarity

It did.

Oh look! I found a match:

Jim says, "As I understand, only the portion of the federal debt that is owed to the public is included in the FRED Total Credit Market Debt."

In other words, the portion of the Federal debt held "privately" by the Federal government itself, is not included in TCMDO.

Back in April, when Gene Hayward caught a discrepancy in my debt numbers, he wrote,"I teach my students that the total US Debt (14.3T) is comprised of a Public portion and a Private portion. In your chart am assuming you are only including the public portion with the private subtracted in creating the 'public debt' line."

So both Gene and Jim recognize that the big number ($14.3T) includes debt held by the public, plus debt held internally by the government. And that the smaller number ($9.6T or whatever) includes only the debt held by the public.

Okay. It's startin' to sink in.


jim said...

Every debt is a liability for one and an asset for another. The debt that a govt or a corporation or a bank owes to itself
internally cancel each other and are not regarded as part of the credit market. Whether looking at the total or just the part that is owed by one sector, the FRED is only accounting for that which is owed by one sector to other sectors.

The main thing to notice in the debt owed by every one except the federal govt is what is happening to non-federal debt in the last 3 years. The current trend in private sector debt is the entire story of current economic situation. All other factors are dwarfed by comparison to this one factor. In 2008 the private sector went from borrowing at rates that set new records with each successive year to saving at record pace. Total borrowing went from plus 4 trillion in annual new debt created to minus trillions in new borrowing (i.e. repayment and saving replaced borrowing). That is a huge swing and the only time anything like this has happened before is the great depression. Current data shows no period in modern times that the non-federal debt has become less than the year before:

(Annual Change in Total debt)
-(AnnualChange in federal debt)
=(Annual change in private debt)

In economics this is what is generally known as a "liquidity trap".

In a normal functioning economy there are usually more who are willing to borrow than save. Interest rates regulate the relation of borrowing to saving. When interest goes up it encourages people to not borrow and they save instead and when interest rates go down more are encouraged to borrow and less to save.
So normally, if one has $1000 in earnings and spends $900 and saves $100 then the $900 becomes someone else's income and $100 is borrowed and spent thus also becoming income for someone else. In fact, more than $100 is usually borrowed and the next
iteration of income becomes larger than the original $1000 and the economy grows as income grows.

But what happens when there are not enough willing to borrow? If the $900 is spent and $100 saved but no one wants to borrow that $100, then income contracts and in the next iteration there is only $900 of which $810 is spent and $90 is
saved. Easy to see how this excess saving produces a downward spiral in income. And the paradox is it also produce less saving than the savers want to save - this is "the paradox of thrift"

Richard Koo has an excellent presentation on how this dynamic has played out in Japan.

Currently the federal govt is doing most of the borrowing. Much of the private sector is trying to repair their balance sheets and thus private debt is decreasing at unprecedented levels. One of the issues that Koo raises is that the FRED data on how much the private sector is deleveraging are based on estimates that are currently appear to be not very accurate (he calls the data "garbage".
The degree to which income is being diverted to savings and repayment of debt may be understated in the FRED data. That means even more money is being retained in saving and private saving that is detracting from income is larger than the FRED data indicates.

At any rate the federal govt is currently compelled to borrow or face a shrinking economy, because there are too few in the private sector willing to borrow.

The Arthurian said...

Excellent, jim. I even understood the whole thing. Not big on "presentations" and watching video stuff, but I will get to it.

I have been watching "debt per dollar" (total debt relative to M1 money) since the 1970s (Historical Statistics has the numbers for 1916-1970). By 1970, DPD was as high as it ever went during the Depression, so I have long been expecting the collapse shown on your graph.

At any rate the federal govt is currently compelled to borrow or face a shrinking economy, because there are too few in the private sector willing to borrow.

I think it would solve the problem, or help, if we just eliminated existing debt. The Federal Reserve, rather than shredding Treasury securities it holds, should shred the MBSs. Treasury could have a couple of those trillion-dollar coins struck, deposit them at the Fed, and use the funds to make payment for people on existing debt. That would free up income which might then help bring the economy back to life.

Something like that, anyway.