I think there is a tradeoff between inequality and full employment that becomes exacerbated as technological productivity improves. This is driven by the fact that the marginal benefit humans gain from current consumption declines much more rapidly than the benefit we get from retaining claims against an uncertain future.

Wealth is about insurance much more than it is about consumption. As consumers, our requirements are limited. But the curve balls the universe might throw at us are infinite. If you are very wealthy, there is real value in purchasing yet another apartment in yet another country through yet another hopefully-but-not-certainly-trustworthy native intermediary. There is value in squirreling funds away in yet another undocumented account, and not just from avoiding taxes. Revolutions, expropriations, pogroms, these things do happen. These are real risks. Even putting aside such dramatic events, the greater the level of consumption to which you have grown accustomed, the greater the threat of reversion to the mean, unless you plan and squirrel very carefully. Extreme levels of consumption are either the tip of an iceberg or a transient condition. Most of what it means to be wealthy is having insured yourself well.

An important but sad reason why our requirement for wealth-as-insurance is insatiable is because insurance is often a zero-sum game. Consider a libertarian Titanic, whose insufficient number of lifeboat seats will be auctioned to the highest bidder in the event of a catastrophe. On such a boat, a passenger’s material needs might easily be satisfied — how many fancy meals and full-body spa massages can one endure in a day? But despite that, one could never be “rich enough”. Even if one’s wealth is millions of times more than would be required to satisfy every material whim for a lifetime of cruising, when the iceberg cometh, you must either be in a top wealth quantile or die a cold, salty death. The marginal consumption value of passenger wealth declines rapidly, but the marginal insurance value of an extra dollar remains high, because it represents a material advantage in a fierce zero-sum competition. It is not enough to be wealthy, you must be much wealthier than most of your shipmates in order to rest easy. Some individuals may achieve a safe lead, but, in aggregate, demand for wealth will remain high even if every passenger is so rich their consumption desires are fully sated forever.

Our lives are much more like this cruise ship than most of us care to admit. No, we don’t face the risk of drowning in the North Atlantic. But our habits and expectations are constantly under threat because the prerequisites to satisfying them may at any time become rationed by price. Just living in America you (or at least I) feel this palpably. So many of us are fighting for the right to live the kind of life we always thought was “normal”. When there is a drought, the ability to eat what you want becomes rationed by price. If there is drought so terrible that there simply isn’t enough for everyone, the right to live at all may be rationed by price, survival of the wealthiest. Whenever there is risk of overall scarcity, of systemic rather than idiosyncratic catastrophe, there is no possibility of positive-sum mutual-gain insurance. There is only a zero-sum competition for the right to be insured. The very rich live on the very same cruise ship as the very poor, and they understandably want to keep their lifeboat tickets.

If insurance were not so valuable, it would be perfectly possible to have very high levels of inequality and have full employment. The very rich might employ endless varieties of servants to cater to their tiniest whims. They’d get little value from the marginal new employee, but the money they’d lose by paying a salary would have very little value to them, so the new hire could be a good deal. But because of the not-so-diminishing insurance value of wealth, the value of hiring someone to scratch yet another trivial itch eventually declines below the insurance value of holding property or claims. There is a limit to how many people a rich person will employ, directly or indirectly.

In “middle class” societies, wealth is widely distributed and most peoples’ consumption desires are not nearly sated. We constantly trade-off a potential loss of insurance against a gain from consumption, and consumption often wins because we have important, unsatisfied wants. So we employ one another to provide the goods and services we wish to consume. This leads to “full employment” — however many we are, we find ways to please our peers, for which they pay us. They in turn please us for pay. There is a circular flow of claims, accompanied by real activity we call “production”.

In economically polarized societies, this dynamic breaks down. The very wealthy don’t employ everybody, because the marginal consumption value of a new hire falls below the insurance value of retaining wealth. The very poor consume, but only the most basic goods. In low productivity, highly polarized economies, we observe high-flying elites surrounded by populations improvising a subsistence. The wealthy retain their station by corruption, coercion, and extraction while the poor employ themselves and one another in order to satisfy these depredations and still survive. Unemployment is not a problem, exactly, but poverty is. (To be “unemployed” in such a society means not to be idle, but to be laboring for an improvised subsistence rather than working for pay in the service of the elite.)

Idle unemployment is a problem in societies that are highly productive but very unequal. Here basic goods (food, clothing) can be produced efficiently by the wealthy via capital-intensive production processes. The poor do not employ one another, because the necessities they require are produced and sold so cheaply by the rich. The rich are glad to sell to the poor, as long as the poor can come up with property or debt claims or other forms of insurance to offer as payment. [1] The rich produce and “get richer”, but often they don’t much feel richer. They feel like they are running in place, competing desperately to provide all the world’s goods and services in order to match their neighbors’ hoard of financial claims. However many claims they collectively earn, individually they remain locked in a zero-sum competition among peers that leaves most of them forever insecure.

It is the interaction of productivity and inequality that makes societies vulnerable to idle unemployment. The poor in technologically primitive societies hustle to live. In relatively equal, technologically advanced societies, people create plenty of demand for one another’s services. But when productivity and inequality are combined, we get a highly productive elite that cannot provide adequate employment, and a mass of people who preserve more value by remaining idle and cutting consumption than by attempting low-productivity work. (See “rentism” in Peter Frase’s amazing Four Futures.)

One explanation for our recent traumas is that “advanced economies” have cycled from middle-class to polarized societies. We had a kind of Wile E. Coyote moment in 2008, when, collectively, we could no longer deny that much of the debt the “middle class” was generating to fund purchases was, um, iffy. So long as the middle class could borrow, the “masses” could simultaneously pay high-productivity insiders for efficiently produced core goods and pay one another for yoga classes. If you didn’t look at incomes or balance sheets, but only at consumption, we appeared to have a growing middle class economy.

But then it became impossible for ordinary people to fund their consumption by issuing debt, and it became necessary for people to actually pay down debt. The remaining income of the erstwhile middle class was increasingly devoted to efficiently produced basic goods and away from the marginal, lower productivity services that enable full employment. This consumption shift has the effect of increasing inequality, so the dynamic feeds on itself.

We end up in a peculiar situation. There remains technological abundance: “we” are not in any real sense poorer. But, as Izabella Kaminska wonderfully points out, in a zero-sum contest for relative advantage among producers, abundance becomes a threat when it can no longer be sold for high quality claims. Any alternative basis of distribution would undermine the relationship between previously amassed financial claims and useful wealth, and thereby threaten the pecking order over which wealthier people devote their lives to stressing and striving. From the perspective of those near the top of the pecking order, it is better and it is fairer that potential abundance be withheld than that old claims be destroyed or devalued. Even schemes that preserve the wealth ordering (like Steve Keen’s “modern jubilee“) are unfair, because they would collapse the relative distance between competitors and devalue the insurance embedded in some people’s lead over others.

The zero-sum, positional nature of wealth-as-insurance is one of many reasons why there is no such thing as a “Pareto improvement”. Macroeconomic interventions that would increase real output while condensing wealth dispersion undo the hard-won, “hard-earned” insurance advantage of the wealthy. As polities, we have to trade-off extra consumption by the poor against a loss of insurance for the rich. There are costs and benefits, winners and losers. We face trade-offs between unequal distribution and full employment. If we want to maximize total output, we have to compress the wealth distribution. If inequality continues to grow (and we don’t reinvent some means of fudging unpayable claims), both real output and employment will continue to fall as the poor can serve one another only inefficiently, and the rich won’t deploy their capital to efficiently produce for nothing.

Distribution is the core of the problem we face. I’m tired of arguments about tools. Both monetary and fiscal policy can be used in ways that magnify or diminish existing dispersions of wealth. On the fiscal side, income tax rate reductions tend to magnify wealth and income dispersion while transfers or broadly targeted expenditures diminish it. On the monetary side, inflationary monetary policy diminishes dispersion by transferring wealth from creditors to debtors, while disinflationary policy has the opposite effect. Interventions that diminish wealth and income dispersion are the ones that contribute most directly to employment and total output. But they impose risks on current winners in the race for insurance.

Why did World War II, one of the most destructive events in the history of world, engender an era of near-full employment and broad-based prosperity, both in the US where capital and infrastructure were mostly preserved, and in Europe where resources were obliterated? People have lots of explanations, and I’m sure there’s truth in many of them. But I think an underrated factor is the degree to which the war “reset” the inequalities that had developed over prior decades. Suddenly nearly everyone was poor in much of Europe. In the US, income inequality declined during the war. Military pay and the GI Bill and rationing and war bonds helped shore up the broad public’s balance sheet, reducing indebtedness and overall wealth dispersion. World War II was so large an event, organized and motivated by concerns so far from economic calculation, that squabbles between rich and poor, creditor and debtor, were put aside. The financial effect of the war, in terms of the distribution of claims in the US, was not very different from what would occur under Keen’s jubilee.

Although in a narrow sense, the very wealthy lost some insurance against zero-sum scarcities, the post-war boom made such scarcities less likely. It’s not clear, on net (in the US), that even the very wealthy were “losers”. A priori, it would have been difficult to persuade wealthy people that a loss of relative advantage would be made up after the war by a gain in absolute circumstance for everyone. There is no guarantee, if we tried the jubilee without the gigantic war, that a rising tide would lift even shrinking yachts. But it might very well. That’s a case I think we have to make, before some awful circumstance comes along to force our hand.

[1] It is interesting that even in very unequal, high productivity societies, one rarely sees the very poor reverting to low-tech, low productivity craft production of goods the wealthy can manufacture efficiently. One way or another, the poor in these societies get the basic goods they need to survive, and they mostly don’t do it by spinning their own yarn or employing one another to sew shirts. One might imagine that once people have no money or claims to offer, they’d be as cut off from manufactures as subsistence farmers in a low productivity society. But that isn’t so. Perhaps this is simply a matter of charity: rich people are human and manufactured goods are cheap and useful gifts. Perhaps it is just entropy: in a society that mass produces goods, it would take a lot of work to prevent some degree of diffusion to the poor.

However, another way to think about it is that the poor collectively sell insurance against riot and revolution, which the rich are happy to pay for with modest quantities of efficiently produced goods. “Social insurance” is usually thought of as a safety net that protects the poor from risk. But in very polarized societies, transfer programs provide an insurance benefit to the rich, by ensuring poorer people’s dependence on production processes that only the rich know how to manage. This diminishes the probability the poor will agitate for change, via politics or other means. Inequality may be more stable in technologically advanced countries, where inexpensive goods substitute for the human capital that every third-world slum dweller acquires, the capacity and confidence to improvise and get by with next to nothing.

So, it is to my great discredit that I had not read Kalecki’s Political Aspects of Full Employment (html, pdf) before clicking through from a (characteristically excellent) Chris Dillow post. There is little I have ever said or thought about economics that Kalecki hadn’t said or thought better in this short and very readable essay.

Here is Kalecki describing with preternatural precision the so-called “Great Moderation”, and its limits:

The rate of interest or income tax [might be] reduced in a slump but not increased in the subsequent boom. In this case the boom will last longer, but it must end in a new slump: one reduction in the rate of interest or income tax does not, of course, eliminate the forces which cause cyclical fluctuations in a capitalist economy. In the new slump it will be necessary to reduce the rate of interest or income tax again and so on. Thus in the not too remote future, the rate of interest would have to be negative and income tax would have to be replaced by an income subsidy. The same would arise if it were attempted to maintain full employment by stimulating private investment: the rate of interest and income tax would have to be reduced continuously.

Dude wrote that in 1943.

Let’s check out what FRED has to say about interest rates during the era of the lionized, self-congratulatory central banker:

Yeah, those central bankers with their Taylor Rules and DSGE models were frigging brilliant. New Keynesian monetary policy was, like, totally a science. Who could have predicted that engineering a secular collapse of interest rates and incomes tax rates (matched, of course, by an explosion of debt) might, for a while, moderate business and employment cycles in a manner unusually palatable to business and other elites? Lots of equations were necessary. No one would have guessed that, like, 70 years ago.

The bit I’ve quoted is perhaps the least interesting part of the essay. I’ve chosen to highlight it because I hold a churlish grudge against the “Great Moderation”.

Bloggers say this all the time, but really, if you have not, you should read the whole thing.

I wanted to add a quick follow-up to the previous post, inspired by its very excellent comment thread. (If you do not read the comments, I’ve no idea what you are doing at this site; I am always pwned by brilliant commenters.)

Cribbing Minsky, I defined the core of what a bank does as providing a guarantee. Bill Woolsey and Brito wonder whatever happened to maturity transformation, the traditional account of banks’ purpose? Nemo and Alex ask about interest, which I rather oddly left out of my story.

Banks do a great many things. They certainly do charge interest, as well as a wide variety of fees, both related and unrelated to time. Banks do borrow short and lend long, and so might be expected to bear and be compensated for liquidity, duration, and refinancing risk. Banks also purchase office supplies, manage real estate, and buy advertising spots. Banks do a lot of things.

But none of those are things that banks do uniquely. Banks compete with nonbank finance companies and bond markets for the business of lending at interest, and nearly every sort of firm can and occasionally does borrow short to finance long-lived assets. There is no obvious reason why any special sort of intermediary is needed to mediate exchanges across time of the right to use real resources. As Ashwin Parameswaran points out, there is no great mismatch between individuals’ willingness to save long-term and requirements by households and firms for long-term funds. Banks themselves largely hedge the maturity mismatch in their portfolios, outsourcing much of the whatever risks arise from any aggregate mismatch to other parties. Once upon a time, before we could swap interest rate exposures and sell bonds to pension funds, perhaps there was a special need for banks as maturity transformers. But if that was banks’ raison d’être, we should expect their obsolescence any time now.

But that is not and never was banks’ raison d’être, however conventional that story might be. Banks’ role in enabling transactions is and has always been much more fundamental than their role in lending at interest over long periods of time. It is not for nothing that we sometimes refer to cash money as “bank notes”. In some times and places, paper bills issued by private banks served as the primary means of commercial exchange. Here and now, the volume of exchange of bank IOUs to conduct transactions entirely dwarfs the scale of loans intended to survive for an extended length of time. When we buy and sell with credit and debit cards, merchants pay a fee for nothing more than a banks’ guarantee of a customers’ payment. Banks issue deposits to merchants with varying degrees of immediacy for these purchases, but charge no interest at all to debit and nonrevolving credit card customers.

Interest over time is, of course, at the center of how banks make money. But the question is why banks have any advantage over bond investors and nonbank finance companies in earning an interest spread. One traditional story has to do with relationship banking: local banks know local businesses (in part by having access to the history of their deposit accounts, in part because of social and community connections), and are able to lend profitably and consistently to firms whose creditworthiness other lenders could not evaluate, and can helpfully smooth over time variations in lending interest rates due to changes in the firm’s financial situation over time because the ongoing relationship prevents firms from jumping during periods when they are “overcharged”. I think there is something to this story historically, but fear it’s growing less and less relevant as the business consolidates into megabanks that require “hard” centrally verifiable statistics rather than “soft” local information to justify making credit decisions. In any case, if managing relationships actually is banks’ special advantage, note that it has nothing to do with maturity transformation and everything to do with evaluating creditworthiness and providing a guarantee.

Even if there’s something to the relationship banking story, it’s not sufficient to explain the resilient centrality of banks. Why can’t local nonbank finance companies couldn’t enter into persistent relationships with firms, evaluate creditworthiness, and earn the same smoothed interest rates as banks? Banks’ advantage in earning an interest rate spread comes ultimately not from anything special about their portfolio of assets, but from what is special about their liabilities. Banks pay no interest at all or very low interest rates on a significant fraction of their liabilities, low-balance checkable demand deposits. The class of bank creditors called “depositors” accepts these low rates because 1) they deem the bank to be highly creditworthy, and so don’t demand a credit spread; and 2) they gain an in-kind liquidity benefit because “bank deposits” serve as near perfect substitutes for money.

To me, a bank is any entity that can issue liabilities that are widely accepted as near-perfect substitutes for whatever trades as money despite being highly levered. Bill Gates can issue liabilities that will be accepted as near-perfect substitutes for money, but Gates is not a bank: his liabilities are viewed as creditworthy precisely because he is rich. The value of his assets far exceed his liabilities; he is not a highly levered entity. Goldman Sachs, on the other hand, was a bank even before it received its emergency bank charter from the Federal Reserve. Prior to the financial crisis, despite being exorbitantly levered and having no FDIC guarantee, market participants accepted Goldman Sachs’ liabilities as near substitutes for money and were willing to leave “cash” inexpensively in the firms care. The so-called “shadow banks”, the conduits and SIVs and asset-backed securities, were banks before the financial crisis, because their highly rated paper was treated and traded as a close substitute for cash despite the high leverage of the issuing entities. Shadow banks wrapped guarantees around a wide variety of promises that, after a while, we wished they hadn’t.

Maturity transformation, issuing inexpensive liquid promises today in exchange for promises that pay a high rate of interest over a period of time, is one strategy that banks use to exploit the advantages conferred by the resilient moneyness of their liabilities. Issuing guarantees for a fee — swapping their money-like liabilities in exchange for some other party’s less money-like IOUs — is another way that banks exploit their special advantage. These activities are not mutually exclusive: lending at interest over time and bundling the price of the guarantee into a “credit spread” embedded in the interest rate combines these two strategies.

But maturity transformation is really nothing special for banks: depositors’ willingness to hold the liabilities of highly levered banks at low interest means banks can invest at scale in almost anything and earn a better spread than other institutions. Whether that spread comes from bearing maturity risk or credit risk or currency risk or whatever doesn’t matter. What is uniquely the province of banks is their ability to issue, in very large quantities relative to their capital, money-like liabilities in exchange for the illiquid and decidedly unmoneylike promises of other parties, and thereby effectively guarantee those promises. It is this capability that makes banks special and so central in enabling commerce at scale among mistrustful strangers.

A final point about banks as I’ve defined them is that they simultaneously ought not exist and must exist. In financial theory, the interest demanded of an entity by its liability holders should increase in the leverage of the institution. There ought not to be entities that can “lever up” dramatically, often against opaque and illiquid assets, without creditors demanding a large premium to hold its deposits. Yet banks exist, and existed long before deposit insurance immunized some creditors from some of the risks of bank leverage. They existed in spite of financial theory, with the help of marble columns and good salesmanship and perhaps the small assurance that came from knowing that if depositors were ruined, the banker would be too. Under a wide range of political and institutional settings, banks come into the world and gain people’s trust despite their inherent fragility, because humans are susceptible to elevating gods, because commerce requires scalable and trustworthy guarantors. People require commerce enough to collectively hope for the best and overlook the inherent contradiction between leverage and trustworthiness. Nowadays, we mostly rely on the state to overcome this contradiction. But recall that no one thought that AAA tranches of structured vehicles had a state guarantee, yet they were often treated as near-money assets. It was just unthinkable that the rating agencies could be so wrong on so vast a scale. If you think about it, you’ll come up with other examples of entities that became able to act as banks, to issue near-money paper despite high leverage, because failure became conventionally unthinkable, even in the absence of any state guarantee or ability to extract one via the knock-on costs of failure. Our propensity to anoint banks is ultimately a social phenomenon, not a product of economic rationality.

Then the state itself is a special kind of bank. Like any bank, the state is perfectly creditworthy in its own banknotes. But unlike other banks, many states make no promise that their notes should be redeemable for anything in particular. “Currency-issuer” states (as the MMTers put it) are highly creditworthy because they don’t make clear promises they might be ostentatiously forced to break. That the failure of states is conventionally unthinkable lends another layer of resilience to the state-as-bank. Successful states use their capacity to intervene in economies — both gently through good stewardship and roughly via taxation — to ensure that the liabilities they issue remain valuable and liquid in commerce. This capacity to intervene renders states and state-guaranteed banks somewhat more resilient than private banks, although not infinitely so. Ultimately the value of any state’s irredeemable notes depends on its capacity to organize and tax valuable real production.

Addendum: Note that the subsidy to sellers described in the previous post depends specifically on the notion of banking systems organized and guaranteed by the state, rather than the more general definition of banks used in this piece. A person who surrendered some real good or service directly for the AAA paper of some CDO bore risk and probably got screwed. But if the same vendor surrendered a real resource for Citibank deposits which had been issued to the buyer against that same AAA paper as collateral, the seller was protected from the risk engendered by her semi-transaction. (“Semi”, because she surrendered something real, but received nothing real in return, creating a risk of nonreciprcation that did not exist before, see e.g. Graeber). It is the state guarantee of bank IOUs, whether explicit or tacit, that effectively socializes the risk of a sale, rather than merely obscuring it or rendering it conventionally unthinkable.

When a bank makes a loan, does it create money “from thin air“? Are banks merely intermediaries, where “if people are borrowing, other people must be lending“? I consider these sorts of questions less and less helpful. Let’s just understand what a bank loan is, in terms of real resources and risk.

Suppose I go to my local bank and ask for a loan. The bank says yes, and suddenly there is “money in my account” where there was not before. Am I now a “borrower” and the bank a “creditor”?

No. Not at all. The transaction that has occurred is fully symmetrical. It is as accurate to say that the bank is in my debt as it is is to say that I am in debt to the bank. The most important thing one must understand about banking is that “money in the bank” also known as “deposits” are nothing more or less than bank IOUs. When a bank “makes a loan”, all it does is issue some IOUs to a borrower. The borrower, for her part, issues some IOUs to the bank, a promise to repay the loan. A “bank loan” is simply a liability swap: I promise something to you, you promise something of equal value to me. Neither party is in any meaningful sense a creditor or a borrower when a loan is initiated.

Now suppose that after accepting a loan, I “make a purchase” from someone who happens to hold an account at my bank. That person supplies to me some real good or service. In exchange, I transfer to her my “deposits”, my IOUs from the bank. Suddenly, it is meaningful to talk about creditors and debtors. I am surely in somebody’s debt: someone has transferred a real resource to me, and I have done nothing for anyone but mess around with financial accounts. Conversely, the seller is surely a creditor: they have supplied a real service and are owed some real service in exchange. It would be natural to say, therefore, that the seller is the creditor and I, the purchaser, am the debtor, and the bank is just a facilitating intermediary. That is one perspective, a real resources perspective.

But it is an incomplete perspective. Because in fact, the seller would not accept my debt in exchange for the goods and services she supplies. If I wrote her a promise to perform for her some service of equal value in the future (which might include surrendering crisp dollar bills), she would not accept that promise as a means of payment. I circumvent her fear by writing to the bank precisely the promise that the vendor would not accept and having the bank “wrap” my promise beneath its own. The bank’s job is not to “lend” anything in any meaningful sense. The bank is just a bunch of assholes with spreadsheets, it has nothing real that anyone wants to borrow. The bank’s role is to transform questionable promises into sound promises. It is a kind of adapter of promises, or alternatively, a guarantor. [*]

Let’s suppose for a moment that the bank’s promises are in fact ironclad. With 100% certainty, the bank will deliver to holders of its IOUs the capacity to purchase real goods and services as valuable as those that were surrendered to acquire the IOUs. Now, when I transfer to a seller IOUs the bank has issued to me in exchange for my somewhat sketchy promises, is it still meaningful to refer to the seller as a “creditor”? After all, she has already received something that is unquestionably as valuable as the goods and services she has surrendered. So her situation is flat, “even-steven”. In exchange for her real resources, she has “money in the bank” whose purchasing power is guaranteed. She bears no risk. But the bank still bears the risk that I will fail to honor my promise while it will be required to honor its own without fail. Which is the creditor then, the seller or the bank? It becomes a matter if definition.

So let’s define. A party that has supplied real goods and services in exchange for a promise of future reciprocation is a “creditor from a real resources perspective”. A “creditor from a risk perspective” is a party that bears the risk that a transfer of resources will fail to be reciprocated. When I have taken a bank loan and “spent the money”, the seller becomes a creditor from a real resources perspective, while the bank becomes a creditor from a risk perspective. The role of creditor becomes bifurcated.

More accurately, the role of creditor becomes “multifurcated”. It cannot be true that “the bank” is a creditor from a risk perspective. Remember: the bank is just a bunch of assholes with spreadsheets, it has nothing real that anyone wants. The risk that we claim sits with “the bank” must in fact fall on people who control or have controlled or might control real resources. We need to consider the “incidence” of the bank’s risk. It might be the case, for example, that the bank’s IOUs, its “deposits”, are in fact not solid at all, such that if I fail to repay the loan, the bank will fail to make good on its promises to people who supplied real goods and services in exchange for bank IOUs. In this case, the “creditors from a risk perspective” are the depositors, people who have delivered real goods and services in exchange for promises from the bank. When depositors are on the hook — and only when depositors are on the hook — there is no divergence of identity between creditors from a risk perspective and creditors from a real resources perspective. Only when depositors absorb losses it is fair to describe a bank as a mere intermediary between groups of borrowers and lenders, perhaps performing credit analysis and pooling risk to facilitate transactions, but otherwise just a pass-thru entity.

That is not how modern banking systems work. Bank depositors are almost entirely protected. The actual incidence of bank risk is, um, complicated. In theory, the risk of loan nonpayment falls first on bank shareholders, then on bank bondholders, then on uninsured depositors, and then on the complicated skein of taxpayers and other-bank stakeholders who back a deposit insurance fund, and then finally on holders of inflation-susceptible liabilities (which include bank depositors). In practice, we have learned that this not-so-simple account of the incidence of bank risk is inadequate, it cannot be relied upon, that the incidence of bank risk will in extremis be determined ex post and ad hoc by a political process which favors some claimants over others when the promises that a bank has guaranteed prove less than valuable.

This may all sound confusing, but one thing should be absolutely clear. Under existing institutions, there is little coincidence in the roles of creditor from a real resource perspective and creditor from a risk perspective. Our banks are machines that permit vendors to surrender real resources in exchange for promises the risks of which they do not bear. The risks associated with those promises do not go away. They may be mitigated to some degree by diversification and pooling. They might be modest, in the counterfactual that banks were devoted to careful credit analysis. But these risks must be borne by someone. The function and (I would argue) purpose of a banking system is to sever the socially useful practice of production-and-exchange-for-promises from the individually costly requirement of assuming the risk that promises will be broken, in order to encourage the former. The essence of modern banking is a redistribution of costs and risks away from people who disproportionately surrender real resources in exchange for promises. Under the most positive spin, modern banking systems engineer an opaque subsidy to those who produce and surrender more real resources than they acquire and consume by externalizing and ultimately socializing the costs and risks of holding questionable claims.

Unfortunately, there are a lot of ways of acquiring protected claims on banks besides producing and surrendering valuable real resources. This divergence of “creditor from a real resources perspective” and “creditor from a risk perspective”, between the party to whom real resources are owed and the party who bears the cost of nonperformance, creates incredible opportunities for those capable of encouraging loans that will be spent carelessly in their direction. Incautiously spent loans are unlikely to be repaid, but the recipients of the money never need to care. Industries like housing and education and of course finance depend heavily on this fact. When industries succeed at encouraging leverage that will be recklessly spent or gambled in their direction, they create certain gains for themselves while shifting risks and costs to borrowers and the general public.

Banks are not financial intermediaries in any simple sense of the word. When they “make a loan”, they serve as guarantors, not creditors. The borrower does not meaningfully become a debtor until the loan is spent. Only then do creditors emerge, but the role of creditor is bifurcated. The people to whom resources are owed are not the same as the people bearing the risk of nonperformance. The question of who actually bears the risk of nonperformance has grown difficult to answer, and concomitantly, incentives among bank decisionmakers for caution in creating that risk have weakened, especially relative to the benefits of cutting themselves in on some share of borrowers’ protected expenditures. This bifurcation of the role of creditor also explains why creditors as a political class are relatively indifferent to the upside of a good economy but extremely intolerant of inflation. A good economy means better higher values and better performance on outstanding loans, but creditors who are owed resources but are absolved from risk do not care about the performance of the loans that have become their assets. Those fluctuations, like fluctuations of the stock market, are somebody else’s problem or somebody else’s gain. Protected claimants, people who are owed money by banks or the state (which is itself a bank), can only lose via inflation. They understandably work within the political system to oppose inflation, which would force them to bear some of the cost of the bad loans whose misexpenditures were, in aggregate, the source of much of their wealth.

[*] Update: JW Mason of the remarkable Slack Wire gently chides that I ought to have attributed this point to Minsky. And indeed I should have! From Stabilizing an Unstable Economy (Nook edition, p. 227):

[E]veryone can create money; the problem is to get it accepted.

…

Banking is not money lending; to lend, a money lender must have money. The fundamental banking activity is accepting, that is, guaranteeing that some party is creditworthy. A bank, by accepting a debt instrument, agrees to make specified payments if the debtor will not or cannot. Such an accepted or endorsed note can then be sold in the open market.