Woodward: The 'Mathematical Flaw'

I want to review a section of

Money and the Federal Reserve System:

Myth and Reality

G. Thomas Woodward
Specialist in Macroeconomics
Economics Division

July 31, 1996 Congressional Research Service Library of Congress
CRS Report for Congress, No. 96-672 E

"Economics Division" of what, the Library of Congress?

Under the heading The "Mathematical Flaw", one reads:

A popular theory about the Fed and money creation in the United States is built around the notion of a "mathematical flaw" inherent in introducing money by means of "lending" as opposed to "spending." This theory starts with the observation that money in the United States (and most other countries) is placed into circulation through the purchase of interest-bearing debt.

To inject money into the economy, the Fed buys federal securities, thereby acquiring an asset that pays interest. In the second round of money creation, banks, S&Ls, and credit unions, through the fractional reserve banking system, earn interest on the loans they hold as a consequence of creating checking account money.

This means that for every dollar of money, there is a corresponding dollar of interest-bearing debt. As a consequence of this arrangement, the argument goes, there is only enough money to pay off the principal of existing debt; there can never be enough to pay the interest that accrues on that principal. If there is to be enough money to handle interest payments in the economy, the theory continues, more borrowing must occur to generate the extra money. Of course, additional borrowing under this arrangement would mean even more interest that cannot be paid out of the existing money supply.

Money is created by lending, not spending. Well, yeah, that's how we create money. I don't get all upset about it, and I don't draw conclusions like "there is only enough money to pay off the principal." There'll be more money tomorrow.

I don't get upset about it because our system usually works. Yes, it gives us business cycles and depression-scale cycles -- but we've had those since finance became an industry. We had them even before the Fed, when we were on the gold standard. It is not the existence of debt that creates problems, but the excessiveness of debt.

What is significant so far is that Woodward uses the word "observation" -- meaning he thinks it a fact, not a notion, that money "is placed into circulation through the purchase of interest-bearing debt." Woodward accepts that as fact.

Just to keep the money supply constant under the system, according to this line of reasoning, debt must grow by the rate of interest. Since the economy grows over time, debt must grow at even a higher rate. As compounding occurs, the result is an explosive growth of debt. Thus, the argument is, policy must actually encourage households and businesses to take on new debt just to keep the money supply from shrinking.

Policy actually does encourage households and businesses to take on new debt, to keep the economy growing. The concern about keeping the money supply from shrinking is something recent, arising among policymakers since the financial crisis of 2008 and the threat of debt deflation.

Hmmm. That supports the theory Woodward wants us to reject.

My Debt-per-Dollar (DPD) graph shows "an explosive growth of debt." It also shows that debt varies, relative to the quantity of money in circulation (M1). The rising DPD peaked during the Great Depression. FDR lowered the DPD but it has been rising again since 1946. And the high level now is associated with the threat of another depression. In my view a high DPD is not merely associated with depression-like conditions, but is the cause of them.

I don't understand "MMT" really well, and a lot of it is irrelevant or worse. But I have no doubt that in their focus on "sectoral balances" leads to a solution to the DPD problem. This alternative, I think, makes "the popular theory" irrelevant, the theory Woodward is trying to shoot down with his so-so argument.

Allowing debt to expand is a problem, these theorists argue, because interest costs are a -- if not the -- principal cause of inflation. When the banks make loans, they charge interest. Interest represents a cost of doing business for borrowers which they pass along to consumers in the prices they charge for goods and services. Hence, it is reasoned, the more interest paid, the higher prices must be.

I agree: Interest costs are the principal cause if inflation. Interest is the "factor cost" of money. When DPD is high, this factor cost intrudes upon Adam Smith's factor costs -- wages and profits and rent -- and competes with them. To the extent that this increases costs without increasing production, it must lead to decline or to cost-push inflation.

Of course, credit-use does increase production. Or, we think it does. Or, it does, as long as the accumulation of debt is not excessive. But when we use credit for everything, when we use credit for money, the cost outweighs the benefit and inflation is the result.

If debt must mushroom over time in order to keep the money supply from shrinking, according to this line of thinking, then the cost of doing business must rise faster each year, and so must prices. In short, it is argued, the money supply process demands that debt grow exponentially. As debt grows as a proportion of total production, so do interest payments. And as interest payments grow relative to the rest of real income, it is claimed, prices must rise faster as well.

Debt DOES grow exponentially. I have shown it here and here and here.

But then, GROWTH is exponential. So... maybe it's no big deal. One must expect to see exponential growth in our economy, in money, in debt. In fact, it is the times that debt does not continue to grow exponentially that we see problems. Times like the present.

I am NOT arguing that we need the exponential growth of debt to continue. I am saying that this concept, that economic growth requires the growth and accumulation of debt, is the concept upon which we base economic policy. Therefore, our economy only grows while exponential increase in debt continues.

It is an important point. We must change the concept upon which we base economic policy. Our assumptions are wrong. We do need debt credit-use for growth. But we don't need to let debt accumulate.

We must use repayment of debt as our primary weapon against inflation.

Debt grows exponentially. Since the end of WWII, so has the quantity of money. These facts ought not be surprising. But now, the DPD divides debt by the quantity of money. Exponential growth, found in both numbers, should cancel itself out of the results. In other words, the effect of growth does not appear in the DPD graph. Therefore I point out that the exponential increase shown by the DPD is a consequence not of growth, but of pro-debt policy.

This dilemma, the proponents argue, is the inherent problem that causes instability in the current banking system -- an instability that the authors believe to be responsible for the business cycle.

Yes. The imbalance between money and debt (clearly visible on my DPD graph) is the problem that causes the depression-scale business cycle, as well as instability both in the banking system and in the overall economy.

Unfortunately, Woodward does not identify "the authors" of this view, and I have not a clue who he is talking about.

Most of those who advance this view believe that to correct the inherent instability in the current monetary system and simultaneously reduce inflation, the system of "debt" money must end. They argue that money must be spent into existence, or at least issued without charging interest.

I do not. I do not say the system of "debt" money must end. What I say is, we have to correct the imbalance between money and debt, the imbalance shown on my graph.

The problem is not that "for every dollar of money, there is a corresponding dollar of interest-bearing debt." The problem is that for every dollar of money, there is $35 of interest-bearing debt. The problem is excessiveness.

The problem is monetary imbalance.

This analysis is deficient on four counts. First, the banking system does not behave as presented above. The payment of interest on debts that arise through the money creation process will neither contract the money supply nor result in the growth of debt relative to the money supply. Second, there is no reason for the money supply to equal the sum of debt and interest. Third, debt is such a common and essential part of an economy, there is always plenty of it available for money creation without any need to encourage the creation of more. (The fourth reason, that interest costs are not the cause of inflation, is discussed in another section).

The crucial error made by the above arguments lies in the proposition that once interest is paid by the government to the Fed, money is "extinguished". If the interest earnings were simply put away into a vault until they were lent out again, the authors would be correct. But in fact, the money is spent back into existence.

The part of the Fed's income used for its own expenses and the dividend paid to member banks is, of course, spent back into existence. The rest the overwhelming majority of all of the income earned by the Federal Reserve -- that which is remitted back to the U.S. Treasury, is also spent. Thus, "lending money" into existence does not mean that debt has to constantly increase to make up for the money that is paid in interest and removed from circulation. It is not removed from circulation; interest payments to the Fed re-enter circulation as they are paid for expenses, as they are paid in dividends, and most significantly as they are paid over to and spent by the Treasury.

The argument has similar problems with its claim that money disappears from circulation as interest is paid back to commercial banks. Like the Fed, commercial banks have expenses. They must pay these out of their earnings -- spending them into existence. They also must pay dividends to their stockholders -- again spending them into existence. Most important among their expenses is interest on their deposits. Whether in the form of explicit interest paid to depositors or implicitly paid as free services (such as check-clearing, balance reporting, etc.), these funds are also spent into existence. Even those sums retained to increase the capital of the bank do not have to be lent, but can be used to purchase expansion of the facilities. There is no requirement in the system that interest earnings must be lent back into existence through new loans.

Woodward does not deal with the the part of the problem that concerns me. He says that money taken from circulation to pay interest is "spent back into existence." Okay. But that doesn't mean it is spent back into circulation.

Also, he equates government money creation with private money creation.

Much of the money received as interest just sits in those interest-bearing accounts where it can earn more interest. On my simple analysis, the money moves out of M1 and into savings, so that savings increases vastly relative to M1 in circulation, over the long haul.

Economists may object to my use of M1 and M2, which are old categories of money. I don't care. They are simple to understand. M1 is spending-money, or money that is circulating. M2 counts both M1 and money in savings. I am looking at money that is circulating, and money that is not circulating but in savings. If you don't want to use M1 and M2, that is fine with me. I don't care about labels. I care about concepts. Redo my DPD graph with your numbers.

Since the amount of dollars represented by the interest payment is returned to the spending stream and the money supply, there is no need for banks to lend continuously a sum equal to the interest payment to keep the money supply constant. Hence, there is no force causing debt to grow continuously relative to the available money supply. The current system is not inherently unstable.

Woodward says: "there is no force causing debt to grow continuously relative to the available money supply." But there is. The force creates an exponential increase in the DPD. The force is policy.

Again, much of the money "represented by the interest payment" is NOT returned to the spending stream. This is the problem. First, the money moves from circulation into savings, decreasing M1. Then, it gets LENT back into circulation, increasing debt. Both of these steps increase the DPD and make the monetary imbalance worse.

Nor is there any reason why there must be enough money outstanding to pay off all outstanding debt. The money needs of the economy are much smaller than an economy's total debt. Money circulates; it gets used repeatedly in the course of a year. Transactions take little time. As soon as money is used in one transaction, it is available for use in another. Consequently, the money stock need only be a fraction of the total transactions that take place in a year.

An economy only needs enough money to complete the transactions that occur in the course of normal business -- not a sum related to total debt. And the total amount of money needed is less than the total value of the transactions because the money is used more than once.

Finally, debt is not created because of a need for money. Every economy - even those without money -- has debt. Debt is a necessity in any modern economy. Indeed, debt pre-dates money in that it exists even in barter economies. It comes in a variety of forms and does not consist exclusively of bank loans. It exists because some people do not consume all that they produce, and are in the position to place some of their goods temporarily in the hands of those who need more goods than they have. Resources are not always in the hands of those who can best employ them. Hence, the lending of resources is common and even necessary for economic progress.

This is such a nice little story... But it neglects the long-term accumulations that lead to monetary imbalance and depression.

Consequently, debt is always present in private affairs. Healthy economies can always be expected to have private debt equal to many times the amount of money that they need. Even as some borrowers repay their loans, still others are ready to borrow. Money creation, therefore, does not drive the creation of debt; the debt is already there regardless of how money is created. It is always there. There is plenty of debt to be used by the banking system for the purpose of money creation with plenty more left over. This is true everywhere there are market systems. Debt does not exist because of a need to create money.

Again, a nice little story, but incomplete. Woodward writes: "Even as some borrowers repay their loans, still others are ready to borrow."

Wouldn't it be nice.

I think we can get it to work that way if we re-think policy and realize we have many policies to encourage the use of credit and the accumulation of debt, but no policy to encourage the repayment of debt.

In short, there is no mathematical flaw. And paying money directly out of the Treasury would have exactly the same economic effect as having the Fed create it by "lending."

There is no mathematical flaw. But there are policy flaws.