Earlier, I made two posts about the immense size of the debt being issued by the US, and the rest of the world. Both raised the question about whether or not the money even existed to fund the debt. After all, for every borrower, there must be a lender. Right? This is a very static view of the problem, and a dynamic view indicates that the money to fund the deficits can exist.

Kiting

Kiting is a form of check fraud. Suppose you have two accounts, but need more money than either account. Assuming the banks make funds available before actually clearing a check, you write a check to yourself from one account, then deposit it and withdraw the funds from the other. Of course, if the check actually clears, your account will be overdrawn, so you turn around and write a check from the second account and deposit it back into the first account. This back and forth can be continued, provided there is a lag between access to deposits and check clearing. If both banks clear all outstanding checks at the same time, the deficit in funds is discovered, and the party ends.

This isn’t exactly the same as having a borrower with no lender. The bank unwittingly loans the kiter the money while it waits for the funds to be deposited. Dirty secret, though: banks loan money first, then check if they actually have it. It sounds bizarre, and it is not how bank lending is described in most textbooks, but it is empirically true.

Enter Fractional Reserve Lending

Suppose I have a thousand dollars. I can use that money to buy a thousand dollars in US treasuries. This funds a thousand dollars of deficit spending. The government then pays me a thousand dollars to pave a road, build a bridge, or repair a levy. I could use that thousand dollars to buy another thousand dollars of treasuries, I could deposit it in the bank, or I could spend it on hookers and booze. If I buy the treasuries, this process can repeat, and I can end up owning as large an amount of US debt as I am willing to work for. If I deposit it in the bank, the bank can hold on to a single Benjamin, have its reserve requirement satisfied, and use the other nine Benjamins to buy nine hundred dollars in treasuries. This process can be repeated, but only so that the total amount of treasuries purchased equals the inverse of the reserve ratio (so, 10x). If I spend the money on hookers and booze, the providers of those goods and services are faced with the same decisions I had. Either way, the same thousand dollars can be recycled many times. Since the government only needs to issue the debt at the same time as it pays out, there need not be a big wait for the cycle to repeat many times.

Sweeps

I stated that the reserve requirement for the banks kept the second choice from resulting in infinite recycling of money. Not all bank accounts are subject to reserve requirements. Checking accounts are… sort of. Sweeps allows money to be transferred from checking accounts, which do have reserve requirements, into accounts that do not have reserve requirements. This is supposedly only valid overnight, with the funds being transfered back into the checking account in the morning. However, as I mentioned before, empirical evidence indicates that banks lend money first, then check their reserves. If they lack the reserves, they can borrow money from the Fed. Fed economists have noted that sweeps have therefore nearly eliminated bank constraints imposed by the reserve requirements. The banks get away with kiting, though not forever, and not with infinite amounts.

Smooth Sailing?

The US government has about $3 trillion in debt to issue this next year. Roughly $2 trillion of that is rollover debt, so there’s some hope that this year’s lenders will become next year’s lenders. The additional trillion has some worried, but government spending can almost always be redirected back into purchasing government debt. There are, of course, a couple hiccups. The first is that it is the banks’ choice to purchase treasuries. However, even if a bank chooses to buy something other than treasuries, that money generally ends up in another bank, so it is just a matter of how many times it must recycle before being converted into treasuries. There’s more government debt than M0, M1, or M2 money supplies, so this clearly hasn’t been much of a problem. Instead, it’s an issue of what interest rate is paid on the debt.

There’s also a tiny problem of people taking money out of banks, causing the whole Ponzi scheme known as fractional reserve lending to collapse, but only a quarter of people are unbanked or underbanked. Hopefully they aren’t the hookers or those providing the booze. Empirically, the vast majority of money circulates in banks.

It’s worth remembering that the US government is not operationally constrained when issuing debt because it issues debt denominated in dollars. The US government has the ability to create dollars out of thin air. It can use newly created dollars to buy back its own debt. Instead, the US government is constrained only by the consequences of creating too many dollars out of thin air. Too many dollars, and prices start rising. Worse, there’s a runaway problem of buyers of treasuries demanding more additional dollars in interest than gets printed. As long as the government runs a deficit, it can’t actually solve the debt problem by creating new dollars. The fear that too many new dollars are being printed is responsible for some of the speculation in things like gold, oil, and records. Many would rather have hard assets than dollars (still, the dollars usually end up in a bank, because even those selling hard assets have bank accounts).

In summary, it is not entirely an issue as to whether or not there is enough money to finance government deficits. It is mostly a question of how much such deficits will end up costing.