The balance sheet recession and The Great Stagnation

Many people point out that we are in a balance sheet recession.  I agree with this view but wish to push it one step deeper.  The negative wealth and income effects on debtors are positive wealth and income effects for the creditors.  If the creditors were keener to invest that money in useful, productive activities the economy would be much stronger.  Balance sheet recessions are most problematic when the investment channel is for some reason broken or especially weak.

You might think “Ah, the weak investment channel is due to weak AD.”  And in part it is.  But, if I may quote the Austrians, production takes time and recoveries do not take forever.  Investors are often keen to invest into the swoosh of a future, not too far away, post-recession boom.  But this desire has been much weaker than in many times past.  A lot of the weakness of AD comes from the investment side, and in fact it predated the recession.


Careful, that's NET investment not GROSS investment. It includes depreciation, or "capital consumption" if you like. That's why a pause in business investment will cause the graph to spike negative... your depreciation schedule grinds away, rain or shine.

Long term, gross investment across most categories has been stable, with net investment declining due to shorter depreciation periods. In popular terms, we ain't buildin' things to last nomore! I personally think that's a good thing; others may disagree.

Charts here:

Bob, thanks for the link. Probably the main reason that cap consumption devours so much more of gross investment is that the investment has been going increasingly into shorter-lived equipment and software rather than structures. There are some graphs of that in the PDF chapter at the page you linked to.

Interestingly even if rated by "Investment as a % of GDP" the US ranks #135 of 143 nations based on this list at Wikipedia (circa 2008) . I've never seen any other economic metric where the US ranks so low on a list. (agreed that a high % is not really a suitable target for a mature economy yet being at the bottom of the list may not be ideal either)

It's not that unusual for us to be low on "% of GDP measures" because we're so rich on an absolute basis. Notice other rich countries are also near the bottom; most of the OECD seems to be below 90.

How is the curve negative at some points? What's negative investment?

Curiously, this is the first post I've seen that uses NDP instead of GDP.

As Bob above pointed out, depreciation can make net investment negative.

Ah, household and government are still going down, whereas in previous recessions household and government went up. The personal balance sheets of households, as those of banks, were overleveraged, and deleveraging still has a way to go, so you can understand that, but domestic business turning around to where it was??? Look again: domestic business as a percentage of net investment is one quarter of its median share, although increasing.

That explains the negative investment puzzle. BTW are shorter depreciation periods governed by whether we build to last or IRS rules and accounting wizardry.

The driving force behind shorter useful lives for business investment is the growing
share of information technology (IT) equipment in business equipment. A computer or cell phone has a much shorter useful life than traditional capital equipment or an office building. If you look at real business investment, essentially all the growth since 1980 has been in IT equipment as investment in traditional equipment has stagnated.

Interestingly, if you take this a step further and look at the growth of the net capital stock per employee you see that there was a sharp slow down in that series about 1980-- from 1945 to 1985 the trend growth rate was 1.6%, but since 1975 the trend has been only 1.0%. This is an important explanatory variable in the slowing of productivity growth since 1980.

I don't see how that is the case. As a larger part of the economy becomes services instead of manufacturing, you'd expect investment in non-IT physical capital to decline. The physical capital stock available to any one manufacturing worker has obviously grown faster than 1% since 1975.

There's a tough conceptual problem here. The value in exchange of a computer falls drastically over 3 years, but the value in production does not. The price falls because there are better computers. So do we really want to call this deprection a bad thing?

I don't think there are balance sheet recessions without investment droughts. The drought precedes it which cause falling interest rates and growing debt until the Minsky moment arrives. Exchange rates are a large part of the story but only part. Investment in substitution (wage arbitrage and off shoring) suppresses investment in advancement (research and development). Once there is little more to achieve from the former, it will be forced to return to the latter but the latter is much harder and riskier.

Since the % of government is slightly down since 2008, does this mean that the stimulus package actually didn't have a lot of shovel-ready investment projects? Or is it distorted by drops in local and state investment, while federal government investment spending rose?

Reading this post made me think a bit about AD. Now, I am not an economist, but recalling from my college econ classes, and much of what I read on the internet, AD is the counterpoint of AS. But, in a credit market, is there really AS? Or, to but it more bluntly, do the identities of AD and AS depend on where you stand? As a debtor, you demand credit, therefore you make up AD, and the creditors make up AS. But as a creditor, you demand debt, so you make up an AD for debt, and the debtors make up an AS for debt.

I ask this because much ink/electrons seem to be spilled arguing about the differences in how you treat AD vs AS recessions, but in some circumstances, it seems like both sides of the transaction contain characteristics of both AD and AS, making a distinction between them somewhat meaningless, to my layman mind anyway.

Can someone clear this up for me?

(PS: AD and AS in heavy manufacturing and durable goods - the typical outputs of industry - do appear more stable as to their sides. There's something about the capital requirements of industrial supply that "feels" like it pegs them to AS, but with financials, there seems like no such natural peg.)

How much of the boom and bust 1998-2002 caused by the Y2K computer bubble?

>>>>The negative wealth and income effects on debtors are positive wealth and income effects for the creditors.

Is that strictly and simultaneously the case?
Wouldn't an overleveraged debtor who, for instance, has a large pending loan maturity reduce consumption and investment in advance, either because they want to make a best shot at rolling over the loan (maybe with a modest principal paydown) or out of fear of the effects of bankruptcy, while the creditor counter-party might also reduce consumption and investment because they know that their wealth is probably about to take a negative hit and they are uncertain as to the magnitude? Even when a creditor expects a significant recovery, they might be more cautious due to the uncertain timing of the recovery of the principal.

Also, imagine a debt jubilee where all debts are cancelled. In theory, the creditors' losses equal the debtors' gains. But I'd expect that economic activity in such a world, one where people felt they couldn't rely on "the rules" anymore, could be slow to return to "normal" (whatever that is).

“The negative wealth and income effects on debtors are positive wealth and income effects for the creditors.” Excuse me? When a house loses 20% of its value, that’s a big chunck of wealth lost for the mortgagor (the debtor). But that doesn’t make the creditor one cent better off does it? In fact it makes the creditor marignally WORSE OFF in that the debt of a debtor who is in difficulties may not be worth it’s face value. Anyone fancy buying my Greek bonds at face value? 

Re the fall in investment, the main contributor to that fall would seem from the chart to be a VERY SLOW and long term decline in Domestic Business investment since 1970. That’s the blue line. That slow decline indicates something far more fundamental and long term being at work than just the current recession. It indicates that businesses just don’t need to invest as much as in the 1970s or 80s.

Plus given a slow long term rise or decline in any of the constituents of aggregate demand, there should not be a problem changing other constituents so as to compensate. So I don’t see that this low level of investment contributes to the current “weakness in AD”.


Sorry, See my comments below for the "something far more fundamental." This trend was noted as far back as 1932 in "The Modern Corporation and Private Property" by Berle and Means. The stages of capitalism are discussed by Carroll Quigley in " The Evolution of Civilizations" (1961; Liberty Fund Reprint 1979).

The wages of policies explicitly designed to promote consumption at the expense of savings?

Maybe we've been eating too much of our seed corn in that respect.


Oddly enough, I can't create anything like this using the data from FRED. The two private business investment categories (which overlap) are "Private Non-Residential Fixed Investment" and "Private Non-Residential Fixed Investment: Equipment and Software." Both rise (as a percentage of GDP) until the first quarter of 2008. Software & equipment bottoms out in the first quarter of 2009 and FI bottoms out in the 4th quarter of 2009. Both have risen more rapidly (as a % of GDP) in this recovery than they did in the recovery following the 2001 recession. Net private investment, again, using the data I can find in FRED, remains positive throughout.

Residential fixed investment, which barely budged as a % of GDP in 2001(in fact it rose slightly from 5.1% at the beginning of 2001 to 5.3% by the end of 2003), really did collapse beginning in 2006--falling from 6.2% in the 4th quarter of 2005 to its current low of has not begun to recover yet.

Doing all this as a % of NDP (GDP - depreciation) makes no noticeable difference.

So I'd guess some of this is an artifact of specific data chosen by the Progressive Policy Institute, which, of course, we can't determine from their chart (and I couldn't determine from looking at their website).

Good for you, Don! The simplest answers to strange results--
(a) You made a sign error
(b) You messed up your data accidentally.

Nice post, Tyler. Good points.

Set up a tax break for open source programming and count it as government investment.

problem solved.

I suggest another correlation:

During the same period of roughly 25 years, the return on invested capital has declined at the same rate (smoothed) by about 25%.

Similarly the rate of return on assets has declined (reference may be made to Deloite's "Shift Index-2010").

The aggregated surpluses (retained earnings) are not being re-deployed.

Due to the fragmented (into several levels) of ownership of large enterprises, the determination of the disposition of aggregated surpluses by distribution, redeployment, or otherwise, has passed to a managerial class. One can plausibly assume that the motivations of the managers of large enterprises in those determinations will differ from those made in more directly owned enterprises.

The aggregation of surpluses in excess of requirements for production, or improvements, decreases the rate (a %) of return on assets, and supports a managerial objective of “self-financing;” basically, managerial capitalism, which is replacing financial capitalism.

Comments for this post are closed