Asymptosis

I think (after a lot of effort) that I’ve internalized Nick Rowe’s modeling of this question (follow links from here) pretty well conceptually.  His answer is Yes.

There have been thousands of posts and comments across the blogosphere since Nick took Krugman to task on the issue a couple of weeks ago, and Nick has been remarkably generous with his time in helping people understand his thinking. (A kudos also to Bob Murphy.) And it’s worth pointing out that Krugman hasn’t really responded to the core argument head-on. Feel free to follow the threads.

Here’s Nick’s model in brief, in my words:

Government borrowing/bond issuance today — considering only its costs, not the potential up/downsides of the associated spending — propagates incentives into the future, like waves are propagated when you throw a rock in a pool. Those incentives cause the old people in every period to consume more than the young people. In each future period, parents will eat some of their kids’ lunch.

Each generation consumes the same amount as they would have otherwise (because first you’re young, then you’re old, first you’re a child, then you’re a parent). But if government eventually has to tax to pay back the debt, the young people in that period are forced to consume less over their lifetimes, because they don’t get to eat their kids’ lunch.

(Nick acknowledges the point that Jamie Galbraith and others have been making for years: if the future GDP growth rate is higher than the interest rate, on average, the taxation never needs to happen, so the burden is never imposed.)

Here’s why I haven’t updated my priors much based on this thinking:

1. The wave model of propagated old/young bond-buying/spending patterns, extending to eternity, doesn’t seem plausible to me intuitively. It seems like the rock-in-the-pool ripples would probably flatten as they spread through time — maybe (?), as discussed in various posts and comment threads, because over the generations a certain percentage of parents bequeath their bonds to their children. That leaves us with “we’ll have to tax to pay for it eventually, so somebody will have to consume less,” which is pretty much where we started.

2. I wonder whether a real agent-based dynamic simulation modeling of Nick’s scenario, with continuous time and using differential equations, would give the same results. I’m not enough of a mathie to even guess, but I wouldn’t be surprised to see very different patterns and/or results.

3. The huge majority of government bonds are bought and traded not by people but by institutions (many of which — this seems significant — are licensed to create credit money and associated debt ex nihilo, and use government bonds as debt collateral in that process). Those institutions don’t have generations — birth/death, parents/children — and they don’t consume real goods (much). Again, my intuition tells me that these facts would bring complex dynamic interaction effects into play. While it might suffice to simplify by modeling things “as if” children were buying bonds from their parents, I don’t feel confident that that’s true.

4. The question Krugman was really asking, underlying his “is debt a future burden” locution, was: A. should we be taxing more or less? and B. should we be spending more or less? (Hence, should we be borrowing more or less?) Since I’d expect to see complex interaction effects from any of the four sources/uses choice combinations, isolating the question in this way seems like a questionable analytical technique. It’s kind of (not wanting to offend, can’t resist the word) petifogging. I’m not at all sure it provides useful information when divorced from the relevant context.

5. Semantics: assuming the model’s right, that we’re forcing a future generation of people to pay for their parents’ (our?) extra consumption — but not changing the amount that all future people will consume in toto — should we call that “a burden on future generations”? I’ll leave that to the philosophers.

I may (still) be displaying an inadequate understanding of the model in some points here, but I think there’s enough of merit above to discourage certainty — to question the ultimate utility of the model.

In short, if I was running a business of any size (yes: I have done so), with any decent amount of money on the line, I would 1. not give a huge amount of weight to the results of this model, and/or 2. be asking for a much more sophisticated analysis.

Which (#2) leaves us again where we were, wrestling with many/most of the big questions of growth macro.

So when Nick says “I thought we all had this debt burden stuff sorted out 30 years ago,” I agree. The answer was “maybe.” (Especially given the long-term historical reality of the Galbraithian scenario described above.) And in my mind it still is — with a little more weight on the “burden” side.

Cross-posted at Angry Bear.

As (mis)conceived by most economists, money (which they confute here with currency) emerged as a solution to the time problem of barter economies: my spinach is ready now, but your apples won’t be ripe for months. How can we trade? Answer: you give me money for my spinach, and I give it back to you later for your apples.

That armchair-sourced fairy tale has been resoundingly debunked — that’s not how money (or even currency) emerged, and the Adam Smithian butcher/baker barter-exchange economy has never existed. Credit money — first embodied in tally marks on clay tablets — emerged and was in widespread use a couple of thousand years before coinage was invented (the latter largely to pay soldiers, whose itinerancy makes other methods problematic).

But the notion of barter economies lives on.

The whole system of national accounts (the NIPAs), in fact, was constructed by Simon Kuznets and company in the 30s as if we lived in a barter economy — with money being merely a time-shifting convenience, and with no accounting for financial assets at all. That accounting was only added a decade or so later, with publication of the Fed Flow of Funds accounts.

I’d like to suggest that this barter model for the real economy results in a great deal of confusion — including (especially?) among economists — largely because the NIPAs don’t usefully model the distinction between saving wheat (which can be consumed) and “saving” money (which can’t). By “useful” I mean “conceptually tractable, subject to consideration without logical error.”

I’m asserting that the barter model of the real economy results in lots of confusion and logical error. Viz: careful economic thinkers like Nick Rowe, Scott Sumner, and Andy Harless feeling the need/inclination to write lengthy think-pieces on that nature of “S.” Or the widespread misconception that “saving” (by whatever definition) “creates” “loanable funds.”

I’d even go so far as to say that those barter-model-induced logical errors pervade most thinking about economies and economics, both among economists and among the laity.

However: If you want to see a market that does operate as a barter economy, look no further than the market for financial assets. In this market, you trade your checking-account or money-market deposits for shares of Apple stock. Somebody else makes the opposite trade. The transaction is mutual and (effectively) instantaneous and simultaneous. It’s a barter.

In a very real and very counterintuitive sense, there is no money exchange in the financial markets. There are just barter swaps of financial assets.

A proleptic response to inevitable objections, and a definition of terms:

1. By “financial asset” I mean something that has exchange value — somebody will give you something in exchange for it — but that cannot be consumed (directly or through use or time/natural decay/obsolescence), providing actual human value — “utility” — in the process. (Things that can be so consumed, and do provide human utility — and derive their perceived value from that utility — are real goods/assets.)

2. “Money,” then, would be that exchange value as embodied (or metaphorically “stored”) in financial assets. Money cannot, in fact, even exist except as it is so embodied. Financial assets are money’s sines qua non – the things without which it does not exist.

Those financial assets (and the money they embody) can be tallied — represented — on clay tablets or in electronic accounting systems, in mental accounting ledgers (“You owe me”), or in physically exchangeable representations of those ledger tallies, such as dollar bills or stock certificates.

By this definition, a dollar bill or a deposit in a checking account is a financial asset, just as much as a share of stock or a government bond is. Which means that all exchanges of financial assets are swaps. Trades. Barters.

Exchanges for real goods, however, are different. When you buy or sell a real good, money (embodied in a financial asset) moves from one account to another, and — this is key — doesn’t disappear. It keeps getting passed on, exchanged. Real goods move the other way, and do disappear. You’re trading something that only has exchange value for something that can — will – be consumed. Conceptually: Financial assets travel in circles. Real goods travel in one direction only: from birth to death, production to consumption.

A physical metaphor may help: think of the financial system (including physical currency transactions) as a giant wheel, pushing along a conveyor belt of real goods.

But economists try to think about/model these very different situations as identical — as if “money” were being exchanged for bonds (and implicitly, as if those bonds will eventually be “consumed”). Since (mental) models for barter economies must be structured very differently from models for monetary economies,* modeling both of these as the same is problematic, confusing, and productive of logical errors — and perhaps even just plain wrong.

The gist of this thinking is not new — much of it reflects ideas floated long ago by circuitists, chartalists, modern monetary theorists, and other such (g)ists. But I’m hoping the formulation as presented here may be useful to others, as it is to me, in 1. forming mental/conceptual models of how economies work, and 2. evaluating the models bruited by others — notably the inherent validity of their underlying and often unstated assumptions.

I would point out in particular that as with my discussion here, the accounting-based modeling approach of Wynne Godley (and his predecessors, successors, collaborators, and parallel travelers) begins not where Kuznets did — with the interchange of real goods and services — but with the nuts and bolts of financial accounting. This approach imposes logical constraints on economists’ reasoning, constraints that seem sadly lacking in much barter-economy thinking.

As Godley says in the conclusion to his seminal paper:

In contrast to the standard textbook methodology, which starts by making very strong behavioural assumptions based on no empirical evidence at all (for example regarding the shape and role of an aggregate neo-classical production function), [this methodology suggests that] a different paradigm is indicated in which knowledge is gradually built up by empirical study, within the formidable constraints imposed by double entry accounting.

I’m not saying it’s impossible to think logically and coherently using the NIPA/Flow of Funds economic model, with the (confusing) barter and savings models embodied in the NIPAs. I’m saying it’s very difficult — especially since economists aren’t trained in financial accounting — and that as a result logical failures are widespread.

* Nick Rowe quoting Clower: “Hang on. In a Walrasian economy there is one big market where all n goods are exchanged; in a barter economy there are (n-1)n/2 markets where 2 goods are exchanged; and in a monetary economy there are (n-1) markets where 2 goods are exchanged, one of which is money.”

Cross-posted at Angry Bear.

Repeat as needed to avoid cognitive dissonance.

“Trust” percentage minus “Distrust” percentage:

Via: Chart of the Day: Republicans Don’t Trust Anyone (Except Fox News) | Mother Jones.

Cross-posted at Angry Bear.

I’ve noticed that many others, like me, are puzzled by the mechanics of the Saving=Investment accounting identity. How do household savings get instantly and perfectly intermediated, in a period, into investment spending — the purchase/creation of long-term productive fixed assets?

An Aha! for me: According to Krugman’s textbook, they don’t (click for larger):

First a correction: “The savings[sic]-investment identity is a fact of [the] accounting [methods developed in the 30s by Kuznets and company to model production of goods and services the national economy].”

But that aside.

If people spend less than producers expect in a given year, the producers create too much product, and it builds up their inventories. That’s easy to understand. (It’s easier if you think about the producers instead of the car-dealer intermediaries that Krugman talks about.)

In the NIPA model, that inventory is counted as “investment.” This makes sense as far as it goes — that inventory is stored real value, stuff that can be consumed/sold for consumption in future periods. As Mankiw explains it in his textbook:

But this explanation also pretty much obliterates the widespread and sloppy notion that increased saving (“not spending”) results in — causes — more productive “fixed investment.” The increased “investment” resulting from increased personal savings just expands inventory. The causations/incentives driving fixed investment are utterly other.

This also makes sense: when people are spending less (are “saving” more), does that spur producers to invest more in their businesses — to buy/create more fixed assets?

Both recent and immemorial history suggest quite the opposite.

Cross-posted at Angry Bear.

 

 

I don’t usually link to Paul Krugman because everyone reads him anyway, right? He doesn’t need my google juice.

But I have to make an exception here because he adds to my trove of graphs demonstrating how America today — after thirty years of Reaganomics policies that were supposed to be all about freedom, liberty, and economic opportunity for all (yeah, and I have a bridge for sale), is at the bottom of the heap when it comes to economic opportunity.

The shining city on the hill keeps getting smaller and richer, and the slopes leading up to it steeper, rougher, and slicker.

That opportunity is best displayed through intergenerational mobility — what my friend Steve calls “convection.” What are the odds that a child will be in a different economic stratum from his parents? It’s a darned good measure of “meritocracy.”

Here’s the key graphic:

 

On the vertical axis is the intergenerational elasticity of income — how much a 1 percent rise in your father’s income affects your expected income; the higher this number, the lower is social mobility.

As you can see, it’s only getting worse. My explanation is here.

The Great Gatsby Curve – NYTimes.com.

Cross-posted at Angry Bear.

John Scalzi:

…in Ayn Rand’s world, a man who self-righteously instigates the collapse of society, thereby inevitably killing millions if not billions of people, is portrayed as a messiah figure rather than as a genocidal prick, which is what he’d be anywhere else. Yes, he’s a genocidal prick with excellent engineering skills. Good for him. He’s still a genocidal prick.

What I Think About Atlas Shrugged – Whatever.

Cross-posted at Angry Bear.

Imagine that over the next week (in a closed American economy — the rest of the world has never existed) everyone sold all their financial assets, paid off all their debts, and deposited the remaining money (and any currency they have) in their checking accounts. No money-market funds, even. Just banks with reserve accounts at the Fed, holding everybody’s money in “cash.”

All those other financial asset prices would dive to zero. Late sellers would sell for nothing.

Would the remaining money in all the bank accounts equal U.S. government debt? That seems to be the implication of MMT thinking, because the remaining money only exists because it got spent into existence by the government deficit spending — crediting bank accounts with that fiat, ex nihilo money in the first place.

Net financial assets = gross financial assets = government debt

(If the government had always just deficit-spent instead of borrowing to cover its deficits, “government debt” would be replaced by “cumulative to-date government deficit spending.”)

I ask not just for clarity, but because (as always), I’m struggling with the relationship between fixed assets and financial assets, between saving and investment.

It’s said that the true wealth of the nation — the “national savings” — consists of its real assets: stuff that can be consumed in the future through use and time/natural decay. The NIPAs only count “fixed assets” — hardware (equipment), software, and structures, so let’s pretend that those constitute all real assets (which they don’t in actuality — not by a long shot). Net investment — purchases/creation minus consumption of fixed assets — increases the stock of fixed assets/”savings.”

In theory, financial assets are just financialized, monetized representatives, proxies, for the real, fixed assets that underly them. And indeed over the (very?) long term, the quantity of fixed assets and net financial assets rise together. Both are much larger today in the U.S. than they are in Thailand, or the U.S. in 1910. Financial-asset values wander all over — even over decades — based on “animal spirits,” but again in the long term…

If that’s so, then in our thought experiment:

Net financial assets = gross financial assets = government debt = fixed assets

The quantity of fixed assets increases over time through net investment. But by MMT thinking, net financial assets can only increase through government deficit spending (or trade surpluses). What is the mechanism whereby government deficit spending is translated into more net financial assets that embody the increased stock of fixed assets?

I imagine a necessarily political mechanism something like the following:

1. People and businesses buy/create fixed assets, resulting in more economic activity — creating/consuming, buying/selling, spending/income.

2. Those increased quantities (both stocks and flows) create more demand for government services. Both individuals and businesses would be decidedly unhappy, I’m thinking, if today’s government were the same size it was, at least in absolute terms, in 1870. (Conservatives and libertarians may say otherwise, but they’re talking through their hats.)

3. Legislators and executives who don’t provide those increased services don’t get re-elected.

4. Taxation lags behind spending — resulting in deficits — because A) people hate taxes and vote against politicians who raise them, and B) if deficit spending is not sufficient to match the increases in fixed assets, depressions result, and the “fiscally responsible” leaders get voted out.

5. The new money from government deficit spending is used to purchase financial assets, driving their prices up to (roughly) match the value of fixed assets.

This is thinking of government and the Fed as one consolidated entity. If you think of them as separate, you can imagine a different mechanism, in which the Fed and the congress/president are engaged in a constant chicken game over inflation, unemployment, and GDP growth, to determine how and when to increase the amount of money/net financial assets (ultimately through deficit spending) to match the stock of fixed assets.

These mechanics would also explain how buying/creating a bunch of drill presses will — through a long, tangled, and messy political process, and in the long but not the short run — result in more “loanable funds.”

Cr0ss-posted at Angry Bear.

Neil Wilson has done yeoman’s duty to (perhaps) achieve a convergence that has been too-long delayed.

A Double Entry View on the Keen Circuit Model.

Steve Keen is, to my knowledge, the only person who is actually encoding a Godley-esque, MMT-style, accounting-based, stock-flow-consistent dynamic simulation model of how economies work. But many MMTers have been quite hostile or at least resistant to Steve’s work, based on some different concepts of endogenous/exogenous money, and — this may seem trivial but it isn’t, at least as it has played out over time — based on details of single- versus double-entry accounting.

The debate has been quite acrimonious at times, and that acrimony has greatly hindered a convergence that in my eyes would be the most salutary event possible in the development of economic thinking and practice.

You can read the details in Neil’s post, but in short he’s re-jiggered Steve’s accounts to make them conform better to (at least Neil’s view of) standard bank-accounting practices. I’m not qualified to evaluate his new formulation, but I am excited to read Neil’s comment on the post, replying to uber-MMTer Scott Fullwiler:

We need to get all this pulled together into a coherent overall model.

Steve’s up for it. I hope you are too.

I’ll just say: I’m very much up for watching it happen.

Also: run don’t walk to read Steve’s Debtwatch Manifesto, posted last week.

Cross-posted at Angry Bear.

Continuing with one of my current hobbyhorses:

Ryan Avent reports on the American Economic Association meeting, with special attention to a presentation by Robert Hall:

Monetary policy: The zero lower bound in our minds | The Economist.

Mr Hall argued that:

A little more inflation would have a hugely beneficial impact on labour markets,

And a reasonable central bank would therefore generate more inflation,

And the Federal Reserve as currently constituted is, in his estimation, very reasonable; therefore

The Federal Reserve must not be able to influence the inflation rate.

… Why is Mr Hall—why are so many economists—willing to conclude that the Fed is helpless rather than just excessively cautious? I don’t get it; it seems to me that very smart economists have all but concluded that the Fed’s unwillingness to allow inflation to rise is the primary cause of sustained, high unemployment. …a macro challenge that actually boils down to the political economy constraints (or intellectual constraints) facing the central bank.

Emphasis mine.

I, of course, am less charitable, and impute other motives.

Hat tip to David Beckworth, whose feelings I fully understand:

I found the whole affair so depressing that I wasn’t able to drag myself to many sessions.

Cross-posted at Angry Bear.

I want to thank all the commenters on my last post — at Angry Bear, at Asymptosis, and at Mike Norman’s blog. You’ve provided me with exactly the education I hoped to achieve. Here’s hoping others benefited similarly.

I asked: what would happen if the the Fed cut the interest rate on reserves from its current .25% to zero. I was not suggesting it should be done. I simply wanted to understand what would happen if that one variable changed.

I want to summarize the conclusions I’ve come to based on all the discussion.

This is me speaking, based on sifting and considering all the responses. I won’t link to all the excellent comments that brought me to this. (Though I do want to highlight the Angry Bear comment by Bad Tux beginning “At 0% IOR”.  It arguably explains things better than I do here, and at significantly less length.)

First, the market monetarists responses. Scott Sumner said (in comments a while back on his blog, which I linked from my original post):

It could be slightly expansionary, or if accompanied by other moves, wildly expansionary.

I’m presuming he says “slightly expansionary” based on the theory Mark Thoma gives us in a November 17 post that Cameron was nice enough to link to on Angry Bear:

it would slightly lower the incentive for banks to hold cash rather than loaning it out, and more loans would help to spur the economy

So banks could lend a quarter point cheaper, or loosen their lending requirements slightly. Assuming there’s some decent amount of demand at lower rates (elasticity of demand is appreciably > 0), or that good borrowers are asking for loans but being turned down cause they’re too risky, this could have a small effect. Scott’s “other moves” presumably include NGDP level targeting by the Fed, but that’s all beyond the contained question I asked here.

James Oswald — who cites himself as a market monetarist but who seems to understand and adhere to much MMT thinking in his other comments and writings — said at Asymptosis:

There is no reason to think they [reserves] would not decrease back to the pre-IOR levels, at least over time, pushing around around 1.4 trillion dollars of high powered money into the economy and triggering significantly higher inflation.

This doesn’t seem to make any sense at all. (And I rather doubt that the Sumners and Beckworths of this world would agree with it.)

In (simplistic) theory, taking IOR nominally negative (the extreme case) would make banks want to instead hold physical currency, with its higher (zero) nominal return. Continuing the simplistic theory, that more-liquid money would be lent and spent more.

But:

1. There are significant costs and management headaches associated with holding currency — trucks, warehouses, security guards, all that rot.

2. It’s completely unclear why banks holding warehouses full of currency would have any incentive effects on borrowing and lending — hence real-economy purchases/velocity. Lenders and real-economy borrowers do their thing because they see valuable risk/return opportunities in the real economy. Changing the form of banks’ holdings will not affect that real-economy reality. Recent history: the QE trades — giving banks reserves in return for bonds — doesn’t seem to have had such an effect, if the massive runup in excess reserves is any testament.

3. Explaining #2: For banks, currency is (see #1) less liquid than reserves. They’re not carrying it in their pockets so they can buy gum at the corner store. They want to make loans; are they going to make them in cash?

4. Even if they did make the move to currency:

A) They couldn’t all do it; there’s not enough currency around.

B) The effect would be to reduce the amount of currency “in circulation” (it’s stuffed under banks’ mattresses), presumably prompting exactly the opposite of what market monetarists suggest:

a. Less real-economy spending/circulation/velocity and

b. Deflation — dollar bills would be harder to come by, so they’d be more valuable relative to real goods

At least in the discussions I’ve been perusing, this “currency” theory of “pushing” “more-liquid” money into circulation doesn’t make any sense. At all.

Market monetarists do seem to at least loosely and implicitly adhere to the (questionable) theory that people and businesses spend more because they hold money in more-liquid form — and they might even confute bank’s incentives and behavior with people’s incentives and behavior at times — but still this currency thinking is probably not a good or widely held market-monetarist theory. In any case it deserves unequivocal debunking.

So: Numerous cogent and convincing commenters agree that taking IOR to zero would have negligible first-order effects on lending and spending. And (invoking authority here) Mark Thoma agrees, in the post cited above:

It probably wouldn’t do much

But – considering the practical, workaday effects on the financial system such as those depicted in the currency fantasy above — Thoma links to an article by Todd Keister from the NY Fed. In short, IOR of zero would break a whole lot of financial entities’ business models. The gang at Mike Norman’s blog point to the problems already facing primary dealers, which could be (greatly?) exacerbated by a drop to zero. StreetEye on Angry Bear says that it would trash the main-street banking model. And etc. Various institutions would die or just withdraw their services/trades from the financial system.

The second- and third-order effects of such eventualities could have profound negative impacts on the real economy.

Which perhaps explains another thing I’ve been wondering about: why did the Fed institute IOR in the first place, and why did it do so when it did?

We can at least give the Fed  credit for understanding the business models of various financial entities. When they saw interest rates heading toward zero, they instituted IOR to prevent the systemic lockup/breakdown described above.

IOW, nothing (much) to see here folks. Move along.

Sorry if I’m so dull that I had to go through all this to figure it out.

Make sense?

Cross-posted at Angry Bear.