The claims about “unsustainable deficits” gained new urgency this week as S&P warned that it was downgrading US federal government debt from stable to negative (see here for recent debate).
This appeared to be a blatantly political move, designed to influence the debate in Washington, adding fuel to the fire to cut budget deficits.
The deficit hysteria has nothing to do with economics, government solvency, or involuntary default. A sovereign government can always make payments as they come due by crediting bank accounts—something recognized by Chairman Bernanke when he said the Fed spends by marking up the size of the reserve accounts of banks.
Similarly Chairman Greenspan said that Social Security can never go broke because government can meet all its obligations by “creating money” (see here).
Instead, sovereign government spending is constrained by budgeting procedure and by Congressionally-imposed debt limits. In other words, by self-imposed constraints rather than by market constraints.
In addition, government needs to be concerned about pressures on inflation and the exchange rate should its spending become excessive. Finally, it should avoid “crowding out” private initiative by moving too many resources to our public sector. However, with massive unemployment and idle plant and equipment, no one can reasonably argue that these dangers are imminent.
Ironically, the ratings agencies recognized long ago that sovereign currency-issuing governments do not really face solvency constraints. A decade ago Moody’s downgraded Japan to Aaa3, generating a sharp reaction from the government. (more here) The raters back-tracked and said they were not rating ability to pay, but rather the prospects for inflation and currency depreciation. After ten more years of running deficits, Japan’s debt-to-GDP ratio is 200%, it borrows at nearly zero interest rates, it makes every payment that comes due, its Yen remains strong, and deflation reigns. In other words, the ratings agencies got it all wrong—as they usually do.
So, as I predicted two days ago, the market reacted to the US government’s credit downgrade with a big “Ho-Hum”.
The Federal Government has been handed a temporary reprieve by Congress: it won’t be shut down just yet. That gives the Democrats and Republicans more time to haggle over which items to cut. The premise is that the government is “running out of money” as President Obama has put it so eloquently in numerous speeches. Let us first examine that claim and then move on to the real subject of debate: Can a sovereign government run out of money?
The answer is easy: No!
Indeed, a sovereign government neither has nor does not have money (see here). The money government uses to spend is created as it spends. That might sound bizarre or even dangerous. But, in fact, on that score it is not so different from any other spender. (see previous discussion)
Can Your Bank Run Out of Money?
Look at it this way. As economists who adopt the (French-Italian) “Circuit” approach have long argued, when a firm wants to spend it approaches a bank . The bank accepts the firm’s IOU (called a loan on the bank’s balance sheet) and creates its own IOU (in the form of a demand deposit). From the firm’s perspective, the loan is its debt and the demand deposit is its asset. The bank “intermediates” because its IOU (the demand deposit) is more widely accepted in payment than is the firm’s IOU.
Of course, the firm is not going to hold the demand deposit since the whole object of borrowing was to spend. The demand deposit will then get shifted to the seller. Now, it is of course possible for the firm to finance its spending by using a sales receipt—a credit to its demand deposit and matched by a debit to the seller’s account.
But at the aggregate level, all the demand deposits were created as the accounting offset to loans. In other words, sales receipts in the form of demand deposits required some previous bank loan. At the aggregate level, bank “money” is created and therefore equal to bank loans—that is where bank money comes from.
Can the bank “run out of money”? No. It neither has, nor does not have money. It creates the money when it makes a loan; and the purpose of this activity is to finance some kind of spending—on goods, services or assets.
Can this money creation be “excessive” in the sense of causing prices to rise? Yes. Can it be “speculative” in the sense that it helps to fuel an asset price bubble? Yes. Can it be “foolish” in the sense that the borrower defaults and the bank ends up holding a worthless IOU? Yes. Can bank lending and thus money creation be constrained by government regulations and supervision? Yes. Finally, can—and should—the bank exercise self-restraint? Yes.
So, just because we say the bank can always create money “out of thin air” by making a loan and creating a demand deposit that does not mean that it should lend “until the cows come home”, or that it does not face regulatory or self-imposed constraints.
Ultimately, good banking practice requires good underwriting—to ensure it does not end up with too many trashy IOUs; and from the macro perspective, government wants to limit bank “money creation” to finance spending in order to prevent inflationary conditions in markets for goods, services and assets.
Is Sovereign Government Different? Users and Issuers of the Curency
Almost everything that has been said above about the finance of the spending of a private firm applies to a government. Government spending occurs simultaneously with a credit to a private bank account—that is to a demand deposit at a bank. The offsetting liability on the government’s books is a credit to the bank’s reserves at the central bank (which is the “private” bank’s asset). The government cannot “run out of money” because the “money” is created when it spends.
I have detailed many times how the government actually does this—following rules for spending that Congress, the Treasury, and the Fed have worked out—and will discuss the details a bit below. But first let us compare government and nongovernment spending through “money creation” in general terms.
Government spending and hence “money creation” has all the same potential drawbacks listed above (most importantly, inflationary consequences when excessive), save one. A private borrower might fail to make payments on loans through no fault of his own. Of course, there are also deadbeat borrowers who choose not to pay. But private firms (and households) need income, or saleable assets, to raise funds to pay their debts. Default is a possibility.
Sovereign government is somewhat different. We usually say that its “income” is tax revenue—a bit different from wages or profits since taxes are at least in some sense discretionary. Further, the government’s potential “customer base” is the whole economy and potentially all economic activity—anything that can be taxed.
However, that really does not get to the more important difference: government is the sovereign issuer of the currency.
A sovereign government cannot be forced into involuntary default—it cannot go bankrupt in its own currency. Let us see why, comparing a sovereign government with the situation of a “user” of the currency.
As my professor Hyman Minsky used to argue, anyone can “create money” in the sense that anyone can issue an IOU to make purchases. In other words, you can spend by issuing an IOU to your bank, with the bank crediting your demand deposit—which you use for the purchase. The “money” is created simultaneously with the spending.
When we talk about a private borrower/spender (household or firm), the bank is concerned with credit-worthiness. There are, indeed, additional constraints facing the bank, including reserve ratios and capital ratios—plus whatever other regulations and oversight government puts on its regulated banks. In practice, reserve ratios do not constrain banks because the development of inter-bank lending markets (called the fed funds market in the US) plus access to the central bank’s discount window ensure that banks can always get reserves—at a price.
Capital ratios can bind, although again in practice the constraint is loose since a bank faced with a good borrower can move assets off the balance sheet, seek additional capital, or use creative accounting to finesse the requirements.
And, as I argued above, growing lending and spending can have consequences at the aggregate level: inflation and currency depreciation should spending be too large relative to capacity. That is why governments use a range of policies to try to constrain lending and spending—monetary and fiscal policy as well as direct limits on bank lending and (in rare cases) wage and price controls.
When government refuses to oversee and regulate private banks, underwriting standards tend to fall—which allows lending and spending to grow quickly, which can have inflationary consequences. But worse, it can lead to a catastrophic financial crisis—as we are witnessing.
What is particularly strange is the way that we treat sovereign government. The treasury’s bank is the central bank—which handles its payments and receipts. The treasury writes checks on its demand deposit at the central bank and moves tax receipts from its accounts at private banks to the central bank when it wants to spend. In the US, the Treasury tries to end each day with a deposit of $50 million at the Fed. In all these respects, the Treasury and Fed relation is much like that between a household or firm at its bank. With one big exception: the credit worthiness of the sovereign issuer of a currency cannot be called into question by financial markets because it can always make payments as they come due.
The Strange Constraints Put on Treasury
We put two constraints on our Federal Government that we do not put on private firms and households:
a) The Treasury cannot issue IOUs to its own bank;
b) Congress imposes a debt limit on Treasury
Amazingly, we do not constrain any household or firm in such a manner. We do not prevent firms or households from issuing IOUs to their banks—indeed, we would argue that such a constraint would be silly. Nor do we directly impose a specific debt limit on households or firms or even state and local governments, because we believe that “markets” can determine how much any nonsovereign entity ought to issue.
But we do impose these limits on the sovereign government that issues the currency. These constraints are adopted on the misguided belief that they will prevent the government from “spending too much”, which would cause inflation and currency depreciation. Hence, it is supposed, we cannot trust Congress and the President to keep spending under control—the budgeting process alone is not a sufficient constraint. They would happily spend so much that we’d quickly become the next Zimbabwe. Thus, we will prevent the Treasury from “borrowing” directly from the Fed since that would result in “printing money”, and if Congress could pay for everything by “printing money” it would approve every pork barrel project constituents dreamed-up. And without a debt limit, Congress would bury government under a crushing debt load that would threaten its solvency.
Let’s examine these constraints in order.
The first constraint means that Treasury can sell its IOUs (bills or bonds) to anyone EXCEPT its own bank. It can sell bonds to households, firms, or private banks but NOT to the Fed (there is a small exception that we need not go into here). So when the Treasury is deficit spending (meaning it needs to write checks for more than its deposit at the Fed), it cannot simply issue an IOU to the Fed. It must instead sell its bills and bonds to private households, firms or banks.
Here’s the problem. To spend, the Treasury must have deposits in its account at the Fed. It does no good to sell its bonds to the private sector, receiving a demand deposit at a private bank—because it cannot write a check on that account. Just as you can only write checks against your account at your bank the Treasury can write checks only on its account at its bank—the Fed. So, for example, it can sell a bond to Bank XYZ and receive credit to an account it holds at Bank XYZ. To spend it needs to transfer the demand deposit to its account at the Fed. This is accomplished by debiting the Treasury’s account at Bank XYZ, and simultaneously debiting that bank’s reserves at the Fed. The Fed then credits the Treasury’s demand deposit.
In normal times, banks do not hold excess reserves. (These are not normal times—banks in the aggregate now hold a trillion dollars of excess reserves thanks to “Helicopter Ben’s quantitative easing. We will ignore that for now.) In that case, Bank XYZ would find itself short of reserves after the Treasury transferred its deposit. There are several ways a bank can get the reserves it needs: borrow in the fed funds market, borrow from the Fed at the discount window, or sell bonds to the Fed. Note that if there are no excess reserves in the banking system, turning to the fed funds market will only cause the fed funds rate to rise. This is the signal the Fed responds to—either lending reserves at the discount window or engaging in an open market purchase to relieve the pressure in the fed funds market. Ultimately, the Fed is the source of reserves banks need.
Note that if the Fed lends reserves to banks, we end up in a position in which banks have essentially borrowed reserves from the Fed in order to “lend” to the Treasury (holding government bonds). If on the other hand the Fed buys the bonds in an open market operation, we end up in a position in which the Fed holds the Treasury’s bonds, so has effectively “lent” to the Treasury—but only indirectly because it used Bank XYZ as the intermediary. Recall that all these operations are required because we prevent the Fed from buying the bonds directly from the Treasury, thereby providing the Treasury with the demand deposits it needs to write checks. So it is doubly ironic that this prohibition then requires either that the Fed lend reserves to banks so they can buy the bonds, or that it buy the bonds from the banks.
Now, in normal times it really does not matter that we have adopted such a roundabout method of allowing the Treasury to do what any other spender can do—issue an IOU to its own bank. It all operates smoothly with the Fed using a private bank as intermediary to do what Congress prohibits it from doing directly. That is to say, what prevents the Treasury from spending its way toward Zimbabwe land is that it has a budget that must be approved by Congress and the President. The prohibition on Fed purchases of bonds directly from the Treasury is not a constraint at all. If we got a Congress and President that wanted Zimbabwean inflation, they could produce that result by agreeing on a budget of quadrillions of dollars of spending. So in normal times, we rely on rationality in Washington to constrain spending.
But these are not normal times. For two reasons. First because we are trying out Chairman Bernanke’s pet theory: quantitative easing—which is based on the belief that if you buy up all the earning assets held by banks and stuff them full of excess reserves that pay only 25 basis points, they will decide to lend. No, they won’t. Instead, they buy Treasury bonds, and then sell a portion on to the Fed that buys them through quantitative easing.
Second, we keep bumping up against the self-imposed debt limit imposed on the Federal government—not by markets but by Congress. We need to understand that the overall debt limit is repeatedly approached because we are running persistent deficits. And that is because tax revenue has been destroyed by the economic downturn caused by Wall Street’s excesses. So Congress must repeatedly raise the debt limit so that the Fed, Treasury and private banks can perform their little charade to allow the Treasury to spend the budgeted amounts. With a few brief exceptions, total outstanding Treasury debt has grown since the founding of the nation—with Congress raising the debt limit as required to let Treasury issue the bonds that banks, households, and firms want to buy.
This is usually done as a matter of routine. But the Republicans want to hold the debt limit hostage to politics—following Rahm Emmanuel’s dictum that a “crisis” should never be wasted. They intend to gut the social programs they do not like on the pretense that this will reduce the budget deficits that threaten the US with bankruptcy. In fact, cutting the social programs will not significantly reduce overall spending (because they are too small) and the US government cannot be forced into involuntary bankruptcy. Hence, neither argument follows on from the facts.
Indeed, if the US does default on any of its payment commitments, it will be because Republicans force it to do so—by forcing government to shut down because Congress will not raise the debt limit. That is the nuclear option that party politics run amuck could lead to.
Conclusion: The Only Thing to Fear is Fear Itself
I realize that whenever the actual operating details are made clear, the response always is: OMG if the government can spend simply by “keystrokes” then we are doomed to Zimbabwean inflation and eventual default on debt. Hence, we need to limit government’s ability to spend—and this can be done by preventing it from “borrowing from” the Fed, and setting a debt limit.
In reality, it is Congress that holds the fate of the US in its hands. The budgeting procedures are what keep inflation at bay, and the normal financing “triangular” operation that uses the balance sheets of the Treasury, Fed and private banks ensure the government meets its payment commitments.
Unfortunately, by using the debt limit as its hammer to destroy social programs, Congress is now threatening to disrupt financing—raising a possibility (albeit very small) that government might be forced by politicians to do what markets CANNOT force it to do: default on its commitments.
So, ironically and through the backdoor, the Republicans might actually bring on a Greek-style debt crisis on the argument that they are defending us from a Zimbabwean-style hyperinflation.