(Not) Thinking About Money

It’s tempting to abolish the word entirely

Originally published at Wealth Economics

J.W. Mason offers a bang-up post on economists’ thinking about “money,” how economists have thought and talked about it over decades and centuries. There’s even a class syllabus and reading list from his class for his John Jay MA econ students. It’s a very deep dive. (“Thirteen Ways of Looking at Money,” eek.) Highly recommended.

But as one who’s spent nigh-on two decades struggling and studying to understand all these writers using that key term in different, overlapping, and conflicting ways, I’d recommend starting with simple, clear definitions (plural) for that vexed word. It’s multivalent and polysemous. Start by unpacking that. Then use those definitions as solid touchstones to untangle those writers’ writings and conversations.

Going a step further, I’d recommend favoring one definition/meaning as the default that we can all agree on: fixed-price assets. IOW “cash” broadly understood. Assets whose prices never change — slightly simplified, checking deposits and physical cash. The other meanings are important, valuable, and necessary. Just, if we want to depart the default and discuss them instead, be very explicit that we’re talking about a different thing.

We have a modern shorthand for these fixed-price assets, from the language of monetary aggregates. There are two types: M2 (basically checking deposits plus a trivial amount of physical cash), and MB (the “monetary base”) — banks’ deposit holdings in their accounts at the central bank, a.k.a. “bank reserves.”1 2

Fixed-price assets have a key characteristic: An asset portfolio that only contains money assets will never incur any (nominal) holding gains or losses (for better or for worse).

Here are four definitions of money, all commonly used and all perfectly valid. The advice here is to only use “money” in the third sense, or make very clear that you’re discussing one of the others.

money. n.

1. A technology or methodology based on arbitrary “monetary” units of account, first employed ca. 3000–2500 BCE for tallying up the value of diverse ownership claims (“assets”), and designating prices. (Coins didn’t emerge for another ~2,000 years.) “What’s the summed-up numeric ‘value’ of two sheep plus three baskets of grain?”

2. Wealth: the sum of tallied, accounted, owned assets (see #1). Ask a real-estate tycoon or mutual-fund investor, “How much ‘money’ do you have?”

3. Financial instruments, assets, whose market prices are institutionally hard-pegged to a unit of account.3 Fixed-price assets. (The “price” of a dollar bill is always one dollar.) This is a subset of #2. Mostly checking- and money-market deposit balances, and physical cash. “How much ‘cash’ do you have in your portfolio, on your balance sheet — assets/instruments whose prices never change?”

This satisfies another commonly used definition of money: “stuff you can use to buy stuff.” Sellers and money-transfer institutions accept (require) these fixed-price money assets for purchases of both real goods/services and financial instruments, because they have a fixed price. Recipients know exactly how much they’re getting — at least relative to the unit of account.4

This is also the (unstated) definition of the instruments that are tallied in monetary aggregates like M2 and MB: fixed-price instruments.

4. Coins and currency. Physical accounting tokens that make it easy to transfer fixed-price assets from one owner/balance sheet to another. A subset of #3. They’re convenient proxies for the accounted assets, the kind you can use to buy houses and businesses, pay taxes, etc. “How much money do you have in your pocket?”

Money in the #3 sense is just an asset class, or category. Deposits and cash are financial instruments, assets. Here’s part of the US Household sector’s balance sheet (“currency” here means physical currency):

This is where confusion often arises. John Hicks highlights it in the first sentence of Chapter 13, “Interest and Money,” in his 1946 Value and Capital.

“There are certain kinds of promissory documents, usually not reckoned as securities, but included as types of money itself.”

He’s talking about bank deposits. It’s as if they can’t be, and aren’t, both: money and securities. He corrects that widespread error in the very next phrase: These money securities “in fact fall under the same classification.” They’re (fixed-price) securitiesMoney assets. (For me at least, “types of money itself” is incomprehensible.)

And this leads to my next recommendation: in general try to think and talk about (balance-sheet) assets, and where appropriate money assets, not money.

When thinking about spending, for instance, think of it as transferring assets to other economic units in exchange for newly produced goods and services. People buy fancy cars because they have lots of assets, not because they have lots of (or a high proportion of) money assets in their portfolios. The spending transfer, mechanically, requires money assets, but that’s a side issue. It’s a few mouse clicks to swap some ETF shares for M assets that you can transfer for spending.

Likewise, monetarists’ “money supply.” Supply seems to implies a flow of money, but the term actually refers to the money stock. (Monetarists call it supply because it lets them sneak in misplaced supply- and demand-based analysis, which is mostly incoherent when thinking in terms of stocks.) Since money assets only comprise about 10% of household assets (for instance), 16% of financial assets, replacing “the stock of money” in monetarist thinking with the stock of assets delivers a vastly more comprehensive measure and object for (macro)economic analysis.

And we have a perfectly good (better) vernacular term for the stock of assets. We call it wealth. See “wealth velocity” here, for example.

You’ll find defense in “assets” against being confused by other widespread usages, like “savings” (plural). That term implies some vague, undefined stock of holdings; the label “savings” doesn’t appear in any national accounts. But it can only mean “assets” (or net worth). A prime example of that usage is Ben Bernanke’s “global savings glut.” (Note the plural.) Is he talking about a saving flow, or an excess stock of “savings”? If the latter, is he concerned there are too many assets out there, too much wealth?

These are just examples; there are many others. But thinking in terms of assets has a big overarching value: it roots you in the actual, observed, accounted, empirical flows and stocks that are (or should be) the analytical foundation for economics and economic modeling.

I’m very much not dissing J.W.’s post here. (Much less his upcoming book with Arjun Jayadev, Money and Things, which I can’t wait to read.) I’m only suggesting that thinking in terms of (money) assets, focusing on that #3 definition of money (while also considering the other three, as distinct concepts), may serve as something of a Rosetta Stone, helping to translate the thoroughly tangled and inconsistent terminology in the body of literature that Mason and Jayadev are promising to unpack for us.

1 Warning, misnomer alert; banks’ deposit holdings at CBs, their MB assets, are mostly not ring-fenced “reserves” against anything these days.

2 While they’re both fixed-price “money” assets, M2 and MB are very different in their institutional usage and (hence) economic import. M2 is money for the nonbank private sector; you can only hold M2 in bank accounts (and cash). Banks don’t hold M2 assets except in some temporary, mechanical transaction accounts. MB, on the other hand, is “bank money,” meta-money, which only exists as deposits in banks’ CB accounts. Nonbanks don’t have CB accounts, so they don’t and can’t hold MB deposits/assets.

3 That price-pegging is guaranteed and enforced by multiple private and public institutions, notably including but not limited to deposit insurance for bank accounts. When the $65-billion money market Reserve Fund/Primary Fund (not insured by the FDIC/FSLIC or any private bank-insurance institutions) “broke the buck” on September 15, 2008, only offering 97 cents in commercial-bank deposits for $1 in money-market deposits, the U.S. Treasury stepped in within 48 hours to guarantee and prop up the $1 share price of all money market funds. (The funds paid a fee for this temporary but mandatory insurance, dissolved in September 2009.) In practice, in normal times and even extraordinary ones, $1 in M assets always sells for $1 — by definition here, but more importantly by institutional enforcement. Fixed-price is M assets’ sine qua non — the thing that makes them what they are.

4 Fixed-price M assets are a mechanical necessity for payment and settlement systems. Consider the Automated Clearinghouse system (ACH), Paypal, Venmo, Zelle, Fedwire, plus the Fed when they try to resolve all the banks’ reserve claims against each other in their nightly settlement. Imagine if the assets they transfer were variable-priced, their price and value changing between transfer initiation, and receipt/disbursement/settlement. Run screaming.

5 The only literature I know that’s tried to untangle this vexing issue of flow supply vs. stock supply is Clower and Burshaw’s 1954 “Price Determination in a Stock-Flow Economy.”