Banks are obviously different from households. But I think explaining two key differences goes far towards explaining why “endogenous money” theory — often pooh poohed as either confused or obvious — is important to economic thinking.
The first is a dweeby accounting difference. The other, which arises from that, is very, very real.
1. When the banks lend to households, the banks expand their collective balance sheet. New assets (loans due) and new liabilities (deposits payable).
When households “lend” (deposit their currency in a bank, buy bonds), they do not expand their balance sheets. They just shift the composition of their asset portfolios (currency traded for deposits, deposits traded for bonds).
(Of course if households were lending directly to other households, that would expand household balance sheets. But they don’t, hardly at all.)
Unlike banks, households only expand their balance sheets by borrowing. New liabilities (loans payable) and new assets: houses residable, cars drivable, food eatable, education usable, health livable. The stuff of life.
This points to the other key difference:
2. Households consume their assets. They have to, in order to live. The very necessary business of living constantly pushes their balance sheets towards imbalance — topped up, for most households, only through labor.
Not so banks. They don’t consume their assets. They don’t have to, in fact can’t. Financial assets (claims against real assets) can’t be consumed.
Banks’ assets are never diminished through consumption, or through use, decay, illness, obsolescence, or death. Excepting borrowers’ payoff or default, their assets are eternal and immortal (as are the banks, for all intents and purposes).
Which exposes the whole notion of “patient lenders” and “impatient borrowers” as the wholesale claptrap that it is. When the banks expand their balance sheets by lending, they are not displaying “patience.” They are not “saving” in any sense of the word, and they are certainly not “patiently foregoing current consumption.” When a household displays “impatience,” it is the inevitable and inexorable impatience of life, passing.
Bankers face very little or no personal risk from expanding their balance sheets; it’s just how they make money. Households, quite otherwise.
Do with this reality what you will, but don’t tell me that the “patient savers and lenders” construct describes any real world that any of us lives in, or the incentives of the lenders in that world.
Cross-posted at Angry Bear.
Comments
15 responses to “The Pernicious Myth of “Patient Savers and Lenders””
SR,
I am not sure why you are stressing that banks can’t consume. Is there some point, because banks can and do consume but in national accounts, intermediate consumption is not included in calculating the GDP (ie subtracted out from output).
Example: a bank buys coffee from the coffee shop downstairs in its building and it has consumed coffee.
@Ramanan
The magnitude of consumption by banks is trivial.
@Asymptosis
Well yeah but my point was basic: like whether it is possible or not and things like that not about magnitude.
@Ramanan
My point is about asymmetric incentives and reaction functions. Consumption is a non-consideration for bankers when lending/expanding balance sheets. QTC for households. In reality, banks can’t save and households can’t lend. The suggestion of symmetrical incentives in the patient/impatient construct (as, for instance, modeled in Krugman/Eggerston) is delusional.
@Asymptosis
“In reality, banks can’t save and households can’t lend.”
Careful.
Banks’ undistributed profits is their saving.
Households can of course lend.
@Ramanan
Okay not can’t; don’t (much).
Better:
Bank lending is not saving. Personal saving is not lending.
“Bank lending is not saving. Personal saving is not lending.”
YES but it seems you impute some other meaning to these statements although the sentences by themselves are correct.
I’d say do not change the language national accounts/flow of funds.
For example your definition of saving (from the other post) itself is inaccurate.
Also, you say households don’t lend and things. Check the table S.3.a of the US Flow of Funds. Household financial assets are huge which mean they lend on a large scale.
(In national accounts, lending is used as a flow but you seem to want to communicate something else when saying saving is not lending).
@Ramanan:
You’re right about the definition of Personal Saving.
Disposable income: I ignored government for simplicity. Should have said so.
I said spending instead of consumption spending.
In the NIPAs all household spending is consumption spending (RE imputed to the RE industry, durables counted as consumption [unlike the more reasonable IMAs]).
Neither of these changes the fact that a higher Personal Saving rate does not increase any stock of Savings.
Curious: how would you define Savings? What measure would you look at in the Z.1?
@Asymptosis
I don’t like Savings to be used as a stock but let’s use it here because in ordinary language it is used in this sense.
There are several things about saving versus savings and one has to be careful. I agree with you that saving is a residual defined as disposable income minus consumption but it does not mean that it doesn’t add to a stock of something.
If you look at Godley/Lavoie’s text or just SNA 2008’s nice guide, you’ll see that a net worth of an economic unit rises by the amount of saving in any period.
So NW = NW(-1) + Saving
There is a more dynamic matter which is related to the paradox of thrift and here it depends on many things. So you could have a model where a fall in households’ propensity to consume, while leads to a fall in output (other things remaining the same), the actual saving may either increase or decrease depending on your assumption.
@Ramanan
“in ordinary language it is used in this sense”
Right, and its used almost uniformly vaguely, inconsistently, and without definition, with the almost universal assumption that Personal Saving adds to aggregate Savings (whatever that is).
So okay don’t change language of national accounts, fine, but with everyone in the world using “Savings” and few of them having any idea what they mean by it or how it relates to sectoral gross or net (or “net”) Saving, or Personal Saving — much less agreeing on any of that or themselves using the terms consistently even within a single sentence (cf Frances’s mixed and muddled usages) — I hope you’ll understand why I do go on so.
Net Worth is right where I go for Savings as well. And that Saving properly conceived is change in net worth, per your equation. The IMAs make that pretty easy to see and understand. Not so with NIPA’s Gross Saving gotta hold a bunch of accounting stuff in your head to grok it. These endless discussions (not just by me, by a long shot!) demonstrate that in spades.
Key problem: change in Net Worth is very much dependent on changes in asset values — market revaluation. So “animal spirits” directly affect Saving per the equation above — even if income and expenses are unchanged.
That’s one reason I think the IMAs give a so much better, more comprehensible and useful representation — because they have an explicit revaluation account that’s at best obscure in the NIPAs.
This relates directly to the Twitter…uh…”discussion” I’ve been having with Scott Winship, on the difficulty of “properly” accounting for cap gains in aggregate income (hence saving) measures.
Ideally, IMO, to get a good picture of the real sector (HHs and nonfin biz), firms’ Net Worth (including fin biz) should be imputed to HHs at market value. Ditto changes in market value (so, cap gains and losses whether “realized” or not). That also lets you ignore firms’ undistributed earnings under the assumption that the market has priced those in to firms’ Net Worth.
A change in this HH Net Worth (Savings) measure is real-sector Saving/Dissaving. It has a complicated relationship to the NIPA’s Gross Saving — a measure that the IMAs have no need for, and that I think confuses far more people than it enlightens. I’m not saying change it, I’m saying ignore it!
@Asymptosis
I agree with several things except that I am not sure I agree with your general points.
It is true revaluations complicate the issue but it is not something we need to worry about in our discussion now. But you are right, the equation for the connection between net worth and saving should include revaluations.
I am also not sure why you worry so much about gross and net saving. Let’s work in an approximation where the consumption of fixed capital is set to zero for convenience. Of course there is another place where gross versus net becomes important which is discussion around saving net of investment but i guess you are not worried about it at the moment.
So why bring in this gross versus net distinction?
@Ramanan
This is really helping me.
I want people to be able to hold the following simple equality in their heads and think using it, in “native mode”:
Saving = Income – Expenditure = Change in Net Worth (“Savings”)
And that can only be true if Income incorporates asset revaluations.
Income. That’s the definition I should be arguing with. (I certainly wouldn’t be the first.)
“You can’t define Income that way because that’s not how Income is defined.”
But if we defined it differently, it would be so much easier to think about how economies work!
And of course many people have discussed the possiblity of using “comprehensive income” as a more useful measure.
Gross vs net, I was mostly gesturing wryly towards the two different meanings of net saving — NIPA’s and MMTers’. More confusion for poor fools like me!
For me it’s a lot about distinguishing between real-capital creation and money saving, which are all crammed together in a hard-to-understand way in NIPA Savings — both Gross and NIPA’s Net (after CapCon). It’s the whole reason that JKH’s S= koan seems so obvious yet profound.
So:
Income = Labor Income + Capital Income (Dividends, Interest, Rent, etc.) + Asset Revaluations
Saving = Income – Expenditure = Change in Net Worth (“Savingsâ€)
A perfectly coherent set of accounting identities and labels that make economies much easier to think about than the identities economists fixate on.
Just to piss you off, I’ll add that another identity — MV = PY — makes much more sense if M = Net Worth. Back to my Velocity of Wealth fixation.
“And that can only be true if Income incorporates asset revaluations.”
Asset revaluation is not counted as income until realized. However, as I understand him, Piketty would tax asset revaluations in his wealth tax at the end of period whether realized or not. For example when it is said that Bill Gates or Warren Buffet are worth so much or that their net worth change so much, this is paper gain as yet unrealized and untaxed. Is it feasible to tax unrealized asset revaluation?
“I want people to be able to hold the following simple equality in their heads and think using it, in “native mode”
That’s an ambitious project. “Saving” and “savings” are ambiguous terms in ordinary language. They can be defined technically/operationally for use in specific contexts, but confusion is still likely to arise owing to the ordinary language ambiguity.
Unfortunately, this is a pretty common occurrence in economics that is not limited to “saving” and “savings.” It also a problem affecting “investment” where firm investment has a different meaning than financial investment by households, which is actually a form of saving. But most people ordinarily think of saving v. (financial) investment with saving being money saving and investment being saving in non-money financial assets. In addition, as you mention, RE investment is firm investment whereas housing purchases by households are saving in real assets that increase net worth as the asset appreciates.
Use of the same or similar terms in different contexts tends to conflate the contexts even with they are essentially different.
Probably the best solution that can be hoped for is that people specify a meaning technically/operationally in a context and stick to that meaning in the context. But history shows that authors have difficulty doing this when they choose ordinary language terms and give them a special meaning. Might have been better if economists had chosen different terminology, but it’s probably too late to change that now. Probably the best that can be done is clear up confusion when it pops up.
But this is definitely a good exercise to undertake to clear up one’s own conceptual apparatus.
“MV = PY” where “M” + NW is problematical as Piketty observed since NW is not transparent. How would V be figured?
What MV – PY says is that MV = NGDP. But is Y actually real “income.” And how to estimate Y= MV/P
As far as I can see, MV = PY is nice as an identity, but pretty worthless practically.
So there’s a lot to be desired wrt to conceptual clarity and also collecting and processing data into useful and objective information.
The way it looks to me, there is a lot of figuring in economics that exceeds the tolerance level of measurement. So conclusions are reported in terms of a greater degree of precision that the data or methodology support and therefore the accuracy of conclusions is overstated. That is to say the modeling is over-determinative wrt to data. Which risks GIGO. And we know that success rate of economic forecasting is not that accurate overall and often embarrassingly wrong at turning points in cycles, which is where accuracy is most needed.
@Asymptosis
Good you say it’s helping, but you have to define saving as income less consumption and not income less expenditure.
About your point on asset revaluations. OK the actual equation is
NW = NW(-1) + Saving + Revaluations.
Is it better now?
But do not try to change the definition of saving. I know you point out that different systems: NIPA, SNA, Flow of Funds, IMA have things different here and there. But these are minor things compared to other systems you may want to introduce. True NIPA itself is incomplete and one needs flow of funds on top: but the whole conceptual system is there as one. So NIPA is insufficient is not a critique of the system of national accounts and flow of funds etc. It is only a critique of NIPA.