Bob Murphy has a thought-provoking column at Mises.org looking at whether, as deflationists like Mish Shedlock argue, credit should be considered virtually part of the money supply. It’s an important question, since credit expansion over the last 20-odd years has so been massive that a credit contraction will, according to the deflationists, constrict the money supply even with the Federal Reserve and Treasury running the printing presses full throttle.

Murphy looks at some complicating circumstances, but his basic point is that credit should not expand the money supply since each unit of credit, while it may add to the purchasing power of the debtor, comes at the expense of an equal unit of purchasing power on the part of the creditor. If X borrows $10 from Y to make a purchase, Y has $10 less to spend.

What I think is wrong with this is that credit is a moving target. Murphy talks about credit as a short-term loan. But sometimes it’s not short-term, and in any case, any time difference no matter how small could significantly affect prices. Once a debt is paid off, after all, we no longer speak about credit. “Credit” is only something that exists for the time that there’s a debt. But credit in the modern American economy is, at any given time, massive. That’s because some debt is long term and other debt is serial or rolling — you pay off your credit card (or perhaps only part of your credit card) one day only to use it again the next. And just as the debtor doesn’t pay all of his debts instantly (which would make debt merely a substitute for existing money), the credit card company or other creditor isn’t paying its vendors immediately either. Vast amounts of debt float from moment to moment. That debt ordinarily adds to purchasing power (thus acting like an infusion of new money). When people either a.) default on the floating debt, or b.) retire it and don’t take out more debt, aggregate purchasing power diminishes. This is a deflationary pressure.

In assessing how much money-like purchasing power is in an economy, the question that must be asked is not “how much purchasing power would there be if credit/debts were canceled out” but “how much purchasing power is there given that credit/debts are not canceled out at a given time.” The time matters, because even if debt X ceases to exist at time Y, that doesn’t balance things out if a new, larger debt Y comes into existence at time Y. The effect, so long as credit grows, is to increase purchasing power. When credit shrinks, the opposite is true.

Look at any given time-slice from the perspective of retail. Let’s say the only retailers in the world are A and B. Debtors like to shop with their credit cards at A. The creditor — the credit card company — likes to buy products from B. At any given point in time, so long as there is debt in the economy, both A and B can make sales: the credit card company can purchase from B at exactly the same time that the card holder buys from A. The cardholder’s credit does not, at a given point in time, come at the expense of the credit-card company’s cash. If it did, it wouldn’t be credit. So at 11:59 a.m. on Tuesday, both A and B can make sales — real goods are sold, but one purchase is bought with real money, the other is bought with credit. Eventually, the credit-card customer will have to obtain real money to pay off his debt, but even if he does that, he can always rack up new debts. The card-issuer must also come up with real money to pay A as well as B — since the card company has to pay vendors of goods to cardholders — but the company has some time-gap in which to do that. This means, in effect, that retailers or vendors the “true” source of loans — giving up real value now (goods) in the hope of getting paid (by credit-card companies or whomever else) later. That might sound lousy for retail, since the credit-card company gets the interest on the short-term loan to the cardholder while the retailer only gets made whole (and in fact, must pay various fees to the card companies). But the retailer gets to sell more goods than he otherwise could, so as long as he does eventually get made whole, he still benefits.

Conversely, of course, when credit contracts, retailers can’t sell as many goods as they had expected; now they can only sell as many goods as regular money can buy at a given time (assuming, for the sake of simplicity, that all credit were wiped out instantly), rather than as many goods as regular money plus credit can buy at once. Instead of A and B both being able to sell goods simultaneously, now only one of them can sell. The result is a retail crash. If credit were not (virtually) new money, if it were just substituting for other dollars, one would not expect to see an overall cratering of retail — neither A nor B would have to lose sales.

Since credit is effectively new or extra money at any given moment, though, it follows that when credit disappears, aggregate sales fall. That’s not enough to guarantee deflation, but it’s a deflationary pressure.