A Challenge For Bernanke

With the exception of some minor editorial adjustments, this is a February 2006 version of a paper written as a response to the nomination of Ben Bernanke as Federal Reserve Chairman. It is an adaptation of Flying Blind: How the Fed Lost Control and How to Regain It, an unpublished paper written in 2001 which has proved unexpectedly prescient. Please excuse the fact that it may read in places like an elementary economics textbook, a consequence of its divergence from the current mainstream in both its definition of "money" and money's role in financing economic growth.
-- John Atlee


  1. The Tools and Mechanics of Monetary Management
    1. The definition and function of money
    2. Our fractional-reserve money and banking system
    3. The monetary multipliers
    4. The needed reforms
    5. Managing GDP Growth
  2. The Basic Goals of Monetary Policy
  3. The Need for Systematic Coordination of Monetary and Fiscal Policy
  4. Summary
    Appendix: Additional notes on the 100% reserve option


Sophisticated followers of economic news tend to look beyond the Fed's ostensible interest rate targets, and assume the Fed is controlling the economy's money supply, which in turn controls the economy's growth rate, unemployment rate and inflation rate. That is indeed what the Fed should and could be doing. But since 1980, as Fed Chairman Alan Greenspan humbly admitted, in a little noticed 10/19/2000 speech to the Cato Institute, this is an unwarranted assumption:

"...it is essential that we not be deluded into believing that we have somehow discovered the Rosetta Stone of monetary policy...what specifically constitutes money is a notion that has, so far, eluded our analysis."

That explains why the Fed publishes four alternative definitions of "money" (M1, M2, M3 and MZM) -- none of them functionally appropriate -- and why, for the past quarter century, the Fed has used short-term interest rates, rather than money itself, as the ostensible "transmission mechanism" for controlling the economy -- even though there is only a tenuous relationship between interest rates and economic growth. That's not very "transparent."

Since an essential attribute of a true science is the use of consistent, precise and analytically incisive definitions, it is obvious that U.S. monetary management is still an art, not a science. Greenspan's talented artistry managed to avoid another 1970s-type inflation and 1930s-type depression. But American economic performance since 1973 has been far from ideal, including the worst inflation and worst depression since WW II. Our mainstream macroeconomic theory and policy, and monetary management, are in serious disarray. Moreover, it is inconsistent with the principles of democratic government for the nation's economic welfare to be so dependent on one man's economic intuition.

As Fed chairman, Ben Bernanke will have a unique historic opportunity to transform U.S. monetary management and policy into a more truly scientific system that can also provide a model for other countries and international institutions. His general emphasis on transparency, rules and economic models is basic to the needed reforms. But he also needs to address some key flaws in the conceptual underpinnings. We'll discuss the needed monetary management reforms first because they are simpler and less "political" than the needed policy reforms. When we fix the management tools it will be easier to deal with the policy problems.

I. The Tools and Mechanics of Monetary Management

Ironically, the most needed reforms require a conceptual return to the pre-1980 tools: a clear-cut definition of "money," and a high enough reserve ratio on checking accounts (CAs) to give the Fed effective control of money creation.

A. The definition and function of money

What money is -- Pre-1980, the Fed used a simple medium of payment definition -- the paper and coin currency (cash) and the checking account (CA) [1] money we normally use to buy things and pay bills. This is the essential definition for transparent and functionally efficient monetary analysis and management, and the basic definition used in economics textbooks and reference books.  [2]

What money does: its dual role -- In economic flows, money is the financial "blood" of the economic system that endlessly circulates, from buyers to sellers, as the medium of exchange -- the actual means of payment for goods and services. The stock of money -- cash and CAs  [3] -- is the reservoir of this "blood" that accommodates differences in the amount and timing of inflows and outflows. It's the financial counterpart of the physical inventories of factories and supermarkets.

Separating money from non-money -- Pre-1980, there were two key legal/administrative distinctions between money and non-money financial assets. One was checkability -- whether it could be directly transferred from the depositor's account to a payee by a simple order to the depositor's financial institution to make the transfer. The other was the legal prohibition of interest payments on such transaction accounts. This served as an incentive for individuals and businesses to keep only a working minimum of their financial assets in the form of non-interest-earning money.

The Money Demand Ratio (MDR) -- The clear-cut separation of CAs from savings accounts made it easier for the Fed to determine how much money-inventory stock the economy needs to adequately service the current flow of money payments that make up the Gross Domestic Product (GDP). This monetary stock/flow ratio is as important to monetary analysis as the inventory/sales ratio is to nonfinancial business.  [4]

How the Fed lost its way -- The Fed abandoned the clear-cut interest-rate distinction between money and non-money during the devastating worldwide inflation caused by the huge oil price increases of 1973 and 1979. When interest rates on non-money financial assets rose into double digits and bank depositors clamored for similar interest on their CAs, the Fed gave in and permitted banks to pay interest on personal (non-business) CAs under a new name, "Negotiable Order of Withdrawal" (NOW) accounts, and permitted limited checking on the also newly-invented bank and mutual fund "money market" accounts. What the Fed calls M1 is the nearest present counterpart to its pre-1980 definition of money; but M1 is now a hybrid, part medium of payment inventory, part interest-earning savings accounts.

B. Our fractional-reserve money and banking system

Its ancient creation -- Modern CA money was invented by the medieval goldsmiths -- a kind of "reverse alchemy" -- when they realized they could safely lend out many more "deposit receipts" than the amount of gold coins actually deposited with them ("fractional reserve"), and when these deposit receipts came to be used as the normal commercial means of payment (essentially paper money when they needed no endorsement).

The "magical" creation of CA money -- When a bank credits a depositor's CA with a $20,000 loan to buy a car, that particular money didn't exist before the loan -- it didn't come from anyone else's spending, income or saving. It is newly created "out of thin air," as they say, by a sleight-of-hand dual exchange of accounting assets and liabilities during the loan process (a process described in detail in most elementary economics textbooks).  [5]  Functionally, this process is also much like the 19th century banks' creation of printing-press paper currency, but is less transparent.

Money vs. credit -- a basic source of monetary confusion -- In a money-flow perspective, CA money creation, like the minting of metal coins or the printing of paper currency, is a unilateral initiation of a money flow, and the resulting seigniorage profit is a source of income to the creator. In a money-flow perspective, credit is a transaction in which pre-existing money is transferred (advanced) from a lender to a borrower, with the lender expecting its later repayment with interest. The fractional reserve system's linkage of money creation with bank lending -- often called credit creation -- obfuscates this functional flow-initiation by its concept of a dual accounting transaction. The basic money-flow distinction between money and credit may be the monetary "Rosetta Stone" Greenspan is seeking.

The functional role of legally required reserves -- Banks are required to maintain minimum reserves against their depositors' CAs in the form of vault cash (their inventory of currency for cashing depositors' checks) and their own reserve accounts at the Fed. Historically, this was mainly for a cushion against bad loans and recessions. But today they are supposed to provide the Fed with a "lever" for using its "open market operations" to manage the economy.

The transmission of Fed monetary policy to the economy: money growth vs. interest rates -- When the Fed buys Treasury securities from Wall Street bond dealers, this injects Fed-created "high powered money" into the banks' reserve accounts, enabling the banks to create more CA money to lend to business and consumers. When the Fed sells Treasuries, this reduces bank reserves and the banks' ability to create money.

But Fed open market operations also affect interest rates, which, like fungible commodity prices, are largely determined by the balance between supply and demand. When the Fed buys short-term Treasury securities, the money deposited in the bond dealer's bank becomes new reserves that that bank can then lend overnight in the federal funds market to banks that are short on reserves. This increase in the supply of credit in the federal funds market reduces the "fed funds" interest rate. And when banks use these added reserves to create and lend more CA money to business and consumers, this increases the supply of credit -- and reduces interest rates -- in those credit markets. When the Fed sells securities, the opposite occurs.

Before 1980, when the Fed openly managed the money supply, it focused its attention, and the financial press's attention, on the money-growth effects. After 1980, without a clear-cut conceptual or empirical definition of "money," the Fed focused public attention on the interest rate effects of its operations, with the fed funds rate, rather than the money supply growth rate, as the ostensible target of these operations. Thus, when the financial press now reports on this process, they seldom mention the money supply effects, but merely report that the Fed has reduced (or increased) interest rates.

Lower interest rates on home mortgages do leave most home-owning households with more money to spend on other things, and lower business rates reduce the carrying cost of inventories. But for most other aspects and sectors of the economy, including business plant and equipment investment, the functional relationship between interest rates and economic growth is both indirect and difficult to analyze and manage empirically. It is really the other side of the Fed's open market operations, the accompanying change in the money supply -- which the press ignores -- that most affects the economy, and the Fed's focus on the interest rate effect merely tends to obscure this.

That the Fed's present management of the economy's growth rate and unemployment rate is indeed based primarily on the Fed chairman's own talented personal "seat-of-the-pants" intuitive judgment and behind-the-scenes adjustment of bank reserves seems to be "officially" supported by the following annotation in the St. Louis Fed's Monetary Trends chartbook: "[The] Intended Federal Funds Rate is the level...of the federal funds rate that the staff of the FOMC (Federal Open Market Committee) expected to be consistent with the desired degree of pressure on bank reserve positions." Just who actually determines "the desired degree of pressure", and on what basis? Presumably, this is what is decided at the Fed's well-publicized Board meetings, led by the Fed Chairman. But the process is not transparent, and the publicized focus on indirect inflation targets obscures the fact that the actual functional relationship between the federal funds rate and inflation involves the overall economic growth rate and unemployment rate.

Clearly, one of the most needed monetary reforms is to return to the pre-1980 concept of money that can be precisely measured and directly managed to achieve a transparent target of real economic growth. However, there is another reform without which such management is virtually impossible.

C. The monetary multipliers

The Reserve-to-Checkable-Account Multiplier -- Banks are legally required to maintain reserves equal to a certain percentage of their depositors' CA balances. This Reserve/CA Ratio is currently 10%, so each "high powered" dollar of new reserves that the Fed supplies to the banking system theoretically enables the banks to create and lend $10 of new CA money -- i.e., the reciprocal of the 10% Reserve/CA Ratio. However, as we shall see, the actually effective potential multiplier is much greater than that.

The CA-to-GDP Multiplier -- People don't usually put borrowed money in a savings account; they borrow to spend. Thus, in a growing economy, most newly created CA money is spent almost immediately. Moreover, this monetary magic doesn't end with the credit-financed purchase. In the case of a car sale, for example, the dealer takes his profit cut, then replenishes his inventory by buying another car from the car-maker -- who, in turn, pays out income to all those other people who participated in the car's production: workers, parts-makers, bond and stock holders, etc. Each money transaction further increases GDP -- but not by the full amount of the transaction. This is because as the income/spending flow of each successive recipient is increased, they tend to need a larger money-inventory stock in their checkable account balance to service this larger flow -- just as a factory or supermarket tends to need a growing physical inventory to service a growing volume of sales. Therefore, the economic-growth-inducing effect of any given amount of newly created CA money is inherently limited by its gradual absorption into these larger money-inventories. As with the Reserve-to-CA Multiplier, the conceptual value of the CA-to-GDP Multiplier is equal to the reciprocal of the CA MDR (CA/GDP). If the actual CA MDR were a stable 10%, each dollar of new money could finance $10 of additional GDP.

The Reserve-to-GDP Multiplier (henceforth called the Money Multiplier  [6]) -- Combining the Reserve-to-CA Multiplier and CA-to-GDP Multiplier examples gives a theoretical Reserve-to-GDP Multiplier of 100. That would be the Fed's potential "leverage" if it were to try to manage the GDP growth rate by managing the reserves. It would be difficult, but probably not impossible. However, the real reserve to GDP leverage is now far greater than that.

The Vanishing Reserve Ratio -- The traditional CA reserve requirement for the biggest banks used to be 20%. In 1992 the Fed reduced that to 10% for all CAs. In 1994 it permitted overnight "sweeps" (reclassifications) of CAs into savings accounts, which have no reserve requirement at all, thus reducing the reported CA volume and required reserves.

As a result, the banks now have total required reserves of only about $44 billion  [7] -- less than 7% of the $643 billion CA money stock. Moreover, the banks have a normal working vault-cash inventory of nearly $48 billion, which they maintain for cashing depositor checks, regardless of the reserve requirement. In fact, $13 billion of that vault cash is above what they need to satisfy their vault cash reserve requirement. And even that surplus vault cash is 17% more than their $11 billion of reserve balances at the Fed -- which are the only reserves directly affected by the Fed's open market operations. Thus, many, perhaps most, banks are now effectively "unbound" by direct Fed reserve control. And the multiplier of reserve balances at the Fed to GDP (11 to 12600) is a quite unmanageable 1145 to 1. That's like trying to put peas in a bottle with a 20-foot spoon. And that's undoubtedly why Fed reserve management now has to be described in terms of "desired degree of pressure," rather than specific formula-based targets for the growth of reserves, money, real GDP and inflation.

Greenspan's personal economic judgment, based on a vast array of available economic data that is often difficult to integrate and interpret, has usually (but not always) demonstrated a high degree of talent. But managing the economy on the basis of one man's personal judgment isn't scientific, democratic or safe, and it's unfair to leave his successor with that kind of responsibility.

D. The needed reforms

For the Fed to regain adequate control of the money supply -- and the economy -- it mainly needs two basic reforms:

  1. Prohibit interest payments on CAs and prohibit third party checks on interest-paying accounts -- to restore the pre-1980 clear-cut distinction between transaction money inventory and non-money financial assets.  [8]
  2. Sharply increase the reserve requirement on checkable accounts -- at least to the earlier 20%. Indeed, 100% would be much better.

The 100% Reserve Option -- The main historical reasons for reserve requirements (to reduce speculation- and recession-induced bank failures), and for fractional reserves (to permit more flexible money creation better adapted to local economic growth needs) have both been rendered largely obsolete -- by Fed economic macro management and "lender of last resort" role, by the development of an effective national credit market, and by federal deposit insurance. Thoughtful economists have long recognized the potential advantages of a 100% reserve system, but banks have opposed interest-free reserves and the Treasury has opposed paying interest on them.

There is now a unique opportunity for a simple solution to this dilemma, and it may be time to put it on the active policy agenda. The banks now own about $1100 billion of Treasury securities -- far more than needed for balancing their investment portfolios. To meet a 100% reserve requirement, they would merely need to trade less than half of these Treasuries to the Fed, and the Fed would pay interest on their resulting (roughly $550 billion) reserve balances -- a win-win-win-win deal for Banks, Fed, Treasury and economy:

E. Managing GDP Growth

Money growth vs. declining MDR -- Changes in the growth rates of the money stock and the Money Demand Ratio (MDR) both affect the rate of economic growth. An increase in the money supply directly finances an increase in GDP spending. A decrease in the MDR produces a similar result by releasing "excess" money balances for spending. For instance, between 1950 and 1967 the broader MDR (M1/GNP) declined from 40% to 22%, at a relatively stable rate of 3.2% a year, as the rising trend of interest rates made the "opportunity cost" of business and household no-interest CAs seem increasingly expensive, and as the "cash-management" of these money inventories became ever more efficiently computerized and managed by people especially hired for this purpose. This enabled the economy to continue growing with very little long-run growth in the "real" money stock -- but with "cyclical" fluctuations in money growth starkly reflected in the fluctuating GDP growth rates. Since 1967, and particularly since the allowance of interest on CAs, fluctuations in the M1 MDR have complicated the job of monetary management.

The key fomula in monetary management -- GDP is a product of the amount of money available for GDP expenditure, which in turn is influenced positively by increases in the total supply of money (M) and negatively by increases in the tendency to lock money into account balances (the MDR). The relationship between the growth rates of these various factors is expressed in the formula:

GDPgr = Mgr - MDRgr

This basic formula for targeting the policy-desired real GDP growth rate is applicable in either nominal or "real" terms, as long as both money and GDP are in the same terms. And the "M" can be either M1 (including currency) or CA. The Fed needs to carefully monitor the MDR trend growth rate, and manage money growth accordingly -- by precise control of bank reserves. Effectively monitoring the MDR trend requires just as skillful economic analysis as Greenspan's current seat-of-the-pants interest rate management, but is more precise, credible and transparent, and would tend to increase business and financial market confidence.

II. The Basic Goals of Monetary Policy

Inflation vs unemployment -- The "Employment Act of 1946," The Humphrey/Hawkins (H/H) "Full Employment and Balanced Growth Act of 1978," and other economic laws require the Fed to target low inflation and low unemployment. But there has been on-going argument among economists -- and between liberals and conservatives -- regarding which should have priority. Actually, this argument poses a false dichotomy that is politically divisive and economically destructive.

The inherent tradeoff fallacy -- The simplistic Phillips Curve, the related NAIRU (Non-Accelerating-Inflation Rate of Unemployment), and the concept of a "natural" unemployment rate implicitly assume that low unemployment and low inflation are conflicting alternatives, with an inherent tradeoff between them. This misconception pretty much dominated monetary theory and policy during the 1970s and 1980s. It was finally discredited, politically, only by the Clinton/Greenspan achievement of reducing unemployment to the relatively low 4% in 1999-2000 without accelerating inflation. But the underlying errors have not yet been adequately exposed.

The "soft landing" solution -- Neither low unemployment nor high GDP growth rate, in itself, necessarily cause inflation; it's the relationship between them -- as suggested by the familiar concept of a "soft landing" (asymptotic) recovery growth track. It is growing the economy too fast too close to full employment that tends to cause inflationary bottleneck shortages of equipment, inventories, raw materials and particular labor skills. Moreover, a "tight-money" recession, ostensibly aimed at reducing an excessive inflation rate actually disrupts the economy structurally, slowing or stopping the economy's long-run capacity growth, and setting the stage for renewed inflationary bottleneck shortages during the next recovery.

Therefore, a main key to actually achieving the Fed's legal mandate to maintain a stable, full-employment, low-inflation economy is for the Fed to:

  1. explicitly reject its present strategy of using high unemployment as its main anti-inflation tool,
  2. adopt a monetary policy that transparently targets a stable full employment economic environment as its primary goal, and
  3. demonstrate the credibility of that policy by developing an equally transparent systematic money-growth formula for a soft-landing real GDP recovery track to actually achieve that goal.

Then businesses and government agencies can confidently plan for that kind of economic environment in building the needed productive resources -- including educational and skilled labor capital as well as plant and equipment.

Why the primary target of monetary policy should be full employment rather than zero inflation -- The Fed, the Bush Administration, the CBO and the Social Security Trustees are all now "projecting" a continuing average official unemployment rate of between 5% and 5.5% for the indefinite future. (For the Fed, of course, the idea of a "projection is an irony, given that the Fed possesses the ability to bring about any policy target it sets.) Moreover the official unemployment rate does not include the 3% or so of "discouraged workers" who think they have no credible hope of employment -- due to racial prejudice, inadequate education, physical or mental disabilities, place of residence, lack of transportation to jobs, lack of child care, lack of public medical support for relatives with serious medical problems, a legal felony on their record, etc.

During WW II, these "structural" discouraged workers -- and millions more still unemployed from the continuing Great Depression -- were quickly absorbed into the working economy, as workers moved up a rung or two on the skill and salary ladder, as employers (with government help) found creative ways to overcome or offset the disabilities of the discouraged-worker syndrome, and as the army trained those without the skills they needed. The overall unemployment rate quickly fell to close to 1%. Yes, we had wartime price controls and rationing. But that degree of "full employment" was achieved mainly by monetary and fiscal policy tools much the same as those potentially available today -- when we have the will to use them. In fact, if the actual peacetime rate of economic recovery from 1933 to 1940 had continued for five more years without the war, the unemployment rate would have been exactly where it actually was in 1945. The war merely provided the political impetus for responsible use of the macro-management tools that are so badly needed today.

On a soft-landing recovery track towards full employment, a vibrant free enterprise economy encourages both workers and businesses to confidently make needed adaptations to changing opportunities. Between 1992 and 2000, official unemployment declined (almost steadily and at almost "soft landing" rates) from 7.8% to 4%, and the GDP inflation rate stayed quite stable at close to 2%. Responsible macroeconomic policy would enable us to repeat that process and continue to gradually reduce unemployment below that 4%.

Defining "recession" and "depression" -- "Potential" GDP is an estimate of economic capacity based on productivity and labor force factors. Potential GDP is important because, in a growing economy, any period in which GDP is less than Potential should be considered a "depression" -- of greater or lesser degree -- and a "recession" is simply a period in which GDP is receding from Potential, even if in absolute terms it is still increasing.

By definition, Potential GDP includes an assumption of a given level of unemployment. The CBO's estimate of potential GDP assumes 5.2% unemployment. By that definition, the economy is now operating above its potential. I believe this is putting conservative ideology ahead of reality. As a nation, we simply cannot afford a continued effective unemployment rate above 8% for 75 more years, as assumed by the Social Security trustees The social and human rights injustices, the resulting high crime and incarceration rates -- and, "incidentally," the economic and fiscal costs -- are simply too high, with the increased baby-boom retirement rate. More specifically, almost every aspect of Social Security finances is very closely correlated with the unemployment rate. The Social Security Trustees' "low-cost" projection of Social Security finances, which assumes a 4.5% unemployment rate, remains solvent into the indefinite future; the "intermediate-cost" (5.5%) projection does not. A responsible definition of potential should be no more than 4% unemployment.

III. The Need for Systematic Coordination of Monetary and Fiscal Policy

Conceptually, this section uses the CBO's standardized budget, a "dual-deficit" concept which separates the two components of the deficit: the component attributable to Congressional tax and spending policy decisions, and the economic stablization component caused by both automatic and proactive financial responses to high unemployment.

More reserves don't necessarily create more money -- In a severe depression, such as in the 1930s, and the early 1990s in Japan -- and even in less deep depressions, such as in the U.S during 2001-2003 -- private demand for credit (and bank ability to find credit-worthy borrowers) may fall so low that short-term interest rates approach zero. In such an economic environment, Fed efforts to increase money growth (and the resulting money-multiplier effect on GDP spending) by supplying the banks with more reserves is "like pushing on a string," and the banks may have to use their added reserves to buy Treasury securities rather than consumer or business loans.

When banks buy existing securities, that doesn't activate the money multiplier effect unless the seller of the securities quickly spends the proceeds from the sale, which is not inherent in that kind of transaction. More likely, the seller is an investor/speculator who will quickly purchase some other security expected to be a more profitable investment (even if that happens to be, temporarily, a low-interest money market deposit), but in any case will not leave the money in a no-interest CA.

The relationship of federal deficits to bank money creation -- The deficit increases when the Treasury issues new securities. Whether the purchase of these securities by banks increases the CA money supply -- and activates the monetary multiplier -- depends largely on what prompted the increased federal borrowing:

  1. If it is due to a recession-induced increase in the automatic-stabilizer component of the deficit (e.g., increased unemployment insurance, early retirement Social Security benefits, etc.), it will mitigate the effects of the recession but do little to increase the money supply or finance actual recovery.
  2. If the increased deficit is due to a policy tax-rate cut, its effect will be determined by recipients' allocation of the benefits. If it goes mainly to the very wealthy, most will probably be invested in marketable securities (including Treasury securities that financed the tax cut), with any increase in short-term credit supply (e.g. in money market funds) tending to further reduce short-term interest rates. Part will undoubtedly be used to finance an increase in their own consumption spending, Part may also be used for increased real investment spending (depending on profitable investment opportunities in that economic environment). If a policy rate cut goes mainly to middle-class taxpayers, a much larger proportion will undoubtedly be allocated to increased current consumption spending. To replace partisan debate about such allocations with credible empirical data needs more research -- and the Fed has the best resources for this.
  3. If the increased federal deficit is due to increased direct federal spending, or recovery-targeted transfer payments which would be quickly spent, bank purchases of the related Treasury debt securities could potentially produce the same CA increase and monetary-multiplier effect as private consumer or business loans.

To achieve an optimum GDP increase for each dollar of increased deficit, the deficit increase should obviously be well designed for that purpose.

A "Recovery Bonus" fiscal policy tool -- It is far easier to prevent a recession (with the appropriate policy tools in place) than to recover from one that has already caused severe economic structural damage. In either case, direct transfer payments, coordinated with appropriate monetary policy, can get quicker results, and at lower net budgetary cost, than either increased federal spending or tax cuts. And the more "progressive" the transfer benefit (i.e., weighted towards lower income recipients who will spend it most quickly) the more the potential GDP "bang" for each buck of federal deficit that finances it.

One proposal that combines the advantages of many of those offered during the 2000-2003 recession, but has not yet been actually tested, is the recovery bonus. This is a fixed-amount (e.g., $100) weekly (or biweekly) cash payment to every working-age adult and retiree who has a Social Security number, whether employed or unemployed. For those employed, it would be a reduction from the FICA payroll deduction. For others, a check would be sent to their last address of federal record (unemployment insurance, tax records, various low-income payments, etc. The fixed-amount would simplify administration, would be progressive in relation to recipients' incomes, yet would counter the "welfare" designation by also including upper-income FICA payers. Its amount would be automatically reduced as unemployment is reduced. Its budget impact should be combined with the automatic stabilizer component of the deficit, which is similarly reduced by a declining unemployment rate.


The concepts and tools described in this paper:

are all vitally relevant, and useful in the context of Chairman Bernanke's emphasis on transparency, rules, and economic models. The Fed has the wherewithal to further refine these tools and implement the appropriate policies that would make the Chairman's job far easier and less subject to inadvertent setbacks.

These tools have the capacity to create a more dependable, full-employment, low-inflation economic environment than we have yet seen. The transparency inherent in these tools and concepts would make clear to everyone what the Fed's policy goals are. The question then becomes: do the Fed and its chairman have the desire and political will to adopt such goals.

What if the Fed had stopped the credit financing of stock purchases in 1998, before the speculative "bubble" got out of control, and if it had not restricted CA money growth so sharply in 1999-2000. What if a systematically coordinated fiscal policy had slowed the extraordinary trend towards ever-higher surpluses after the unified budget was first balanced 1998 (preferably by meeting some of the long-postponed domestic spending needs)? Could the mid-2000 to mid-2003 recession have been prevented? And could the mid-1992 to mid-2000 gradual reduction of unemployment have been continued for another two years until it reached 3%? We'll never know for sure, but we missed a great opportunity to test it. If Fed Chairman Bernanke adopts the 1995-2000 recovery track model, and President Bush and Wall Street resist interference from those opposed to a full-employment economy, we could get back to 4% unemployment by 2008 and have another chance to test it.

For more on the dynamic money-flow perspective underlying this proposal, see
The Conceptual and Analytical Framework for Macroeconomic Science and Policy on this site.


Note 1
Checking vs. Checkable -- "Checking" is the Fed's continuing name for interest-free business "transaction" accounts; "checkable" also applies to NOW, money market and other post-1980 interest-bearing accounts that permit limited check-writing privileges.
(back to ref 1)

Note 2
Money misconceptions and confusions:

Credit card payments --We usually think of credit cards as another "means of payment". But when we make a credit-card payment, we are actually asking our credit card bank to grant us that much credit and use the proceeds to electronically pay the vendor. Our actual money transfer comes only when we repay the credit-card bank with our own check.

Debit card payments -- these are essentially electronic check payments -- an order to our bank to transfer the amount of the purchase immediately from our checking account to the account of the vendor.

"He has lots of money" -- When we say that Microsoft's Bill Gates "has lots of money," we don't mean that he necessarily has a large checking account balance, but that he has much wealth -- large financial claims on valuable non-money assets such as debt securities or real estate.

"Store of value" -- Traditional lists of the functions of money included "store of value". That was usually valid when most money was in the form of gold and silver coins. But in a modern money economy, "money" itself is a poor store of value because it earns no interest, its value is diminished by inflation, and it may be stolen -- and because it is so easy now to transfer "funds" (financial assets), electronically, in both directions, between checking accounts and other financial assets, such as money-market accounts, that are almost as "liquid" and also earn interest.

"Relative liquidity" -- When the interest-prohibition wall between money and other financial assets was fatally breached in 1980, the Fed no longer had a clear-cut conceptual and empirical definition of "money" as a distinctive financial asset with the unique medium of payment function. So the Fed, and also, unhappily, mainstream financial theory, fell back on the older, very fuzzy, concept of relative liquidity -- the ease and speed with which a non-money asset can be converted into directly spendable transaction money. So it implicitly began to use M2 as its main definition of money. But in the modern electronic financial world of ATMs and money market accounts, "relative liquidity" is a poor criteria for this separation. Today, the practical liquidity difference between a no-interest checking account and an interest-earning money market account is not very significant. (Before ATMs, the weakness of the liquidity distinction was perhaps best illustrated by the fact that, on Sunday, it was easier to get quick money from a used car dealer than from a savings account.)

The anomalous M2-- Once the Fed had breached the interest-prohibition "wall" between interest-free "money" and non-money credit assets, it had no clear-cut definition of "money." So it fell back on the formerly traditional concept of relative liquidity and decided to use what it calls M2 as its main money concept. But less than 20% of M2 is actual medium-of-payment money by the pre-1980 definition (currency and CAs). The rest is non-money -- saving deposits (including bank money market accounts), "small" (under $100K) time deposits (CDs) and "retail" (under $50K) money market mutual funds -- none of which are directly controllable by the Fed, and none has a clear-cut causal relationship to overall economic growth. Moreover, as noted earlier, even M2's M1 component, with its large proportion of interest paying NOW accounts, is now a hybrid concept, part transaction-money inventory, part savings accounts. MZM is a modified version of M2 that still has most of M2's conceptual and empirical disadvantages. M3 is M2 plus large denomination time deposits, money market funds and Eurodollar deposits. No wonder Greenspan was confused about the definition of money
(back to ref 2)

Note 3
Currency -- This paper analyzes only the CA component of the M1 money stock. The currency component -- the amount of currency outside banks -- is actually larger than the checkable accounts (CAs), which are about 5% of GDP. But a far larger dollar volume of transactions is paid by check, and for large (legal) payments a check is usually more acceptable than currency. The ratio of currency to GDP tends to be much more stable than the CA ratio, both in long-run trend and "cyclically," except for national crises like Y2K and 9/11. The currency amount is controlled solely by how much of their total money inventory currency holders want to keep in that form, rather than in their bank accounts. Actually, much U.S. currency is used in other countries as local medium of payment, and by international drug dealers, and various U.S. government agencies. The Fed can neither control its quantity nor even directly measure it; it can only stand ready to meet sharp temporary demands for more during times of financial or other public panic. Since currency is a relatively unimportant factor for most monetary analysis and policy, I don't discuss it in detail in this paper.
(back to ref 3)

Note 4
Money Demand Ratio (MDR) -- Actually, the Fed then used, and still uses, the reciprocal of the MDR, the traditional "velocity" flow/stock ratio (GDP/M). Conceptually, "velocity" is comparable to the nonfinancial factory or store "inventory turnover" ratio. But the functional significance of the MDR stock/flow ("inventory/sales") concept is more transparent, and has greater managerial utility, than the "velocity" concept.
(back to ref 4)

Note 5
The checking deposit created in this transaction becomes an asset for the borrower/depositor, a "credit liability" for the bank; the "mortgage note" on the car becomes a credit asset for the bank, a debt liability for the borrower/depositor. There is no transfer of pre-existing actual money from the bank to the borrower/depositor.
(back to ref 5)

Note 6
The money multiplier -- Keynesian theory stresses the investment multiplier -- the way in which an "exogenous" business investment expenditure (or fiscal deficit) can result in a much greater total GDP increase. But the "exogeneity" of such an investment expenditure or fiscal deficit can depend very much on whether it is "accommodated" by an "exogenous" Fed increase in bank reserves or a reduction in the CA MDR (Keynesian reduction in "liquidity preference").

The role of Checkable Accounts (CA) -- By contrast, the "exogenous" creation of CA money largely finances GDP growth, and the Fed could potentially manage the CA growth rate quite accurately by its open market control of bank reserves. Thus, it is probably more useful, in analyzing and managing monetary relationships, to focus on CA money, rather than the total M1, including currency -- although more research needs to be done on this issue.
(back to ref 6)

Note 7
Unless stated otherwise, empirical amounts cited here for GDP and money factors refer to 2005 3rd quarter values.
(back to ref 7)

Note 8
It would not be necessary to prohibit regularly scheduled automatic transfer or "sweep" arrangements between interest-paying and checkable accounts if the minimum time between such transfers is long enough to maintain effective functional separation between the accounts, as determined by empirical research. (For example, overnight and over-weekend reclassifications would be prohibited.)
(back to ref 8)

Appendix: Additional notes on the 100% reserve option

Key effects of 100% reserve system --

  1. All new money would be created ("printed") directly and transparently by Fed open market operations, rather than "magically" by individual banks.
  2. The fractional reserve (reserve-to-money-growth) multiplier would be eliminated, and the remaining CA-to-GDP multiplier would be a much more precisely manageable ratio of about 25 ($12600b to about $550b).
  3. The 100% reserve requirement applies only to CAs. Banks would still get most of their loanable funds (and new reserves) from the same places they do now -- depositors' interest-paying savings and money market accounts, CDs, and bond and stock issues

This basic reform, long recommended by some eminent economists, would, of course, end the traditional goldsmith-invented fractional reserve system. Most banks have adamantly opposed this because they view interest-free reserve requirements as an "opportunity tax" on them (ignoring the fact that the Fed actually gives most of these reserves to the banking system free of charge). The answer to this political dilemma is obvious and simple: let the Fed pay the banks interest on their reserve deposits at the Fed, at rates comparable to what they now earn on their holdings.

A traditional argument for 100% reserves is that the fractional reserve system, by delegating to private banks the federal government's constitutional right and responsibility to "coin money" (and thereby benefit from the resulting "seigniorage"), is a non-transparent federal subsidy of the banking system. But the traditional complex process by which this is done obscures it so well that even many bankers are honestly quite unaware of what they are doing. The Treasury has traditionally opposed paying the banks interest on their reserves at the Fed because that would reduce the Fed's net interest rebates to the Treasury.

Whatever the moral and/or political justification for these arguments, the 100% reserve reform has not yet generated enough political traction to overcome the opposition -- thus depriving the Fed, and the economy, of the more precise monetary management benefits of a 100% reserve system. It is time to break this historical deadlock by a "practical politics" compromise of interest payment on the reserves.

Finally, since there would no longer be any risk of "runs" on banks due to worry about the security of CAs, banks' would save the cost of FDIC insurance on them.

Written: February 7, 2006
Posted: May 10, 2009
Last revised: October 12, 2010
To IEA home page