Toward the Integration of Economic Science:
The Integrated Dynamic Money-Flow Model (IDMF)

Completing the Keynesian and monetarist revolutions
by integration of concepts and national accounts
in a dynamic money-flow perspective
John Atlee, 4/10/90

 back to  Towards the Integration of Economic Science  main page 


  1. The integrated circular-flow concepts
    1. Credit Market and Primary Credit flows
    2. Monetary Authority
    3. Money stock
    4. Intra-sector money-flows
  2. The IDMF sources-and-uses-of-funds equations
  3. Money vs Credit
    1. The IDMF non-credit concept of money
    2. Money Criteria
      1. Is it actually used as medium of exchange?
      2. Is it actually created in bank lending?
      3. Is its amount actually controlled by the Monetary Agency?
  4. Money as inventory
    1. The money demand ratio (MDR)
    2. The New Money Equivalent (NME) of MDR changes
  5. The Money/Income Multiplier (MIM)
  6. The IDMF equation relating money growth to GNP growth
    1. Forecasting capability
  7. Interest rates
  8. The IDMF structural model
  9. Elastic credit demand
  10. IDMF vs NBER 'business cycle" analysis
  11. Institutional assumptions

Like the Keynesian model, the IDMF model also uses the concept of a circular flow of spending and income. But its primary credit concept makes it possible to fill the financial gap in the Keynesian model with empirical data derived from the FoF accounts. [ 1 ] And its non-credit "pseudocommodity" concept of money makes possible direct empirical analysis of the functional link between money growth and GNP growth.

The integrated circular-flow concepts
(back to contents)

Credit Market and Primary Credit flows

The main macroeconomic money flows are illustrated graphically in Figure 1. The central Credit Market is a conceptual abstraction to represent all intermediate credit flows. The flows in and out are Primary Credit flows, which are comparable to the value added concept in GNP flows. This is the concept which makes it possible to integrate macroeconomic credit flows with national income and spending flows. Primary credit sources are positive financial saving out of current income; uses are the financing of GNP spending.

Conceptually, the Credit Market is an all inclusive "bank" where everyone deposits their financial saving (other than checkable deposits and currency) and from which everyone borrows money to finance GNP spending. (Total inflow and outflow must be equal because of the two-sided nature of credit transactions.) However, the real-world financial institutions that handle these transactions are in the business sector.

The Monetary Authority

The Monetary Authority is another conceptual abstraction. It represents the source of all New Money (increase in money stock). Since, in our economy, new money is created and enters the economy only through bank lending, it is shown conceptually as flowing from the Monetary Authority into the Credit Market. But if we applied the "pseudo-commodity" money concept in practice (e.g., by requiring 100% reserves against all checkable deposits), new money could, if desired, flow directly to the U.S. Treasury as non-credit seigniorage.

Money stock

The circles with "$" signs represent each sector's money stock (checkable deposits and currency).

Intra-sector money-flows

Figures 2, 3, and 4 illustrate the intra-sector structure of money flows. Estimating their empirical values presents a potentially rewarding research challenge.

The IDMF model differs conceptually from both the FoF accounts and Keynesian theory in that it uses a genuine money-flow perspective. Thus, for each transaction, instead of asking only "How does it affect the two parties' static individual balance sheets?" it also asks, "Where toes the money come from, where does it go, and what are the macroeconomic implications?" These are very important questions in the analysis of money creation and various credit flows -- such as those involved in the S&L bailout.

The IDMF sources-and-uses-of-funds equations:
(back to contents)

Income + Primary
+ New
= Spending + Primary
+ Money
I = S ;   PB = PL ;   NM = MBI ;   (4-6)

The first (upper) equation is equally applicable to the economy as a whole and to individual economic units. The other equations are valid only for the whole economy.

Now we can look at some of the IDMF model's conceptual differences in more detail.

Money vs Credit
(back to contents)

The traditional concept of money (to quote the Econ 101 text, with emphasis added) is that money is "essentially the debts of governments and depository institutions." There is thus no clearcut conceptual distinction between money and credit. This way of looking at money treats the process of money creation as merely a static dual (T-account) credit transaction in which the bank's deposit liability is exchanged for the borrower's note, and the bank's note asset is exchanged for the borrower's checkable deposit asset.

In this perspective nothing has actually been created. But in a money-flow perspective this static treatment implies that the bank "borrows" the money which it lends the borrowers from the borrowers themselves essentially all checking account money was -- an obvious absurdity. Thus, the traditional idea initially created by bank loans (or security of "credit creation" is a conceptual anomaly.

The IDMF model makes a precise conceptual distinction between money and credit: money is the inventory stock of the medium of exchange created by the Monetary Authority through the banking system; credit is a transfer of existing money stock to (or receipt of money from) another transactor or sector in a credit transaction, with repayment expected (usually with interest).

When a bank makes a loan to a depositor with newly created money, there is no corresponding transfer of money from the depositor to the bank the money-flow is only one way.

The IDMF non-credit ("pseudo-commodity") concept of money

To clarify the conceptual distinction between money and credit, the IDMF model explicitly rejects the traditional credit concept of money. Although commodity money is obsolete in practice, conceptually integrated macroeconomic analysis must treat money as if it were a costless "commodity" which is produced and quantity controlled by the Monetary Agency. In effect, the "printing press" concept in popular myth is more valid conceptually and analytically than the "credit creation" confusion.

Money Criteria

In the IDMF model there are three specific criteria for distinguishing money from credit in our credit-money economy:

M1 comes closest to meeting all three of these conceptual money criteria. The traditional relative liquidity basis for distinguishing money from other financial assets is functionally inappropriate. The Fed's grouping of M2 and M3 along with M1 as "money and credit aggregates" merely adds to the confusion.

The so-called "money market" is really a short-term credit market. The main demand for credit in this market is not for store-of-value (liquidity") money stock, or even for transaction balances, but rather (as with long-term credit) for credit purchasing power. Similarly, the main supply of credit to this market is from current financial saving rather than a 'liquidity preference" transfer from longer-term investments. (There cannot be a macroeconomic transfer from existing long-term to short-term credit because every seller of such a credit contract must have a corresponding buyer.)

Money as inventory
(back to contents)

Conceptually (and also in practice) money stock performs much the same inventory function in relation to money flows that physical inventory stock performs in relation to production and distribution flows. And in both cases, in order for the stock to adequately service the flow, the nominal value of the stock must increase when the nominal value of the flow increases, whether this increase is due to real economic growth or to inflation. This relationship was emphasized in the 1973-74 and 1979-80 OPEC "oil tax" inflations. In each case the Fed caused a serious recession by restricting the growth of the nominal -- and thus also the real (deflated) -- money supply.

Under relatively normal conditions (i.e., except for war, hyperinflation, financial crisis, illegal activity, etc.) the money balance of a business, government, or household, is mainly working capital which must be financed (either out of income or by borrowing) in the same way as physical inventory. With today's high interest rates neither kind of stock is a very good store of value in the traditional sense. [ 4 ] Adding to money stock constitutes "saving" only in the same technical sense as adding to business inventory or canned goods in pantry shelves (i.e., output not consumed).

The money demand ratio (MDR)

Just as it is important, for both analysis and policy, to monitor current changes in physical inventory/sales ratios, so it is important to monitor current changes in the money/GNP ratio (or, preferably, the ratio of money to adjusted final sales), which the IDMF model refers to as the Money Demand Ratio. Figure 5 shows the long-run trend of this ratio from 1960 to the financial deregulations of the late 1970s.

The MDR is, of course, the reciprocal of the traditional concept of monetary velocity, which is the functional counterpart of physical inventory turnover. But the MDR differs both conceptually and functionally from "velocity." Instead of focusing on the circular flow of money payments, the MDR focuses attention on the money stock and the factors. that influence microeconomic cash management. The GNP denominator is merely an appropriate "scaling device" for systematically relating money stock to the growth-trend of the economy.

The New Money Equivalent (NME) of MDR changes

In relation to money flows, a reduction in the MDR frees money which has been previously locked into the economy's normal-and-necessary money inventory, in much the same way that physical inventory stock can be freed for sale by more efficient computerized inventory management.

When this freed money is spent it becomes an exogenous input to the circular flow of income and spending, thus functioning as an empirically measurable New Money Equivalent of an actual increase in money stock. For instance, from 1948 to 1981 the NME financed most of our real economic growth, with very little increase in the real money stock.

Superficially, changes in the MDR and NME have causes and effects similar to changes in the Keynesian liquidity preference schedule. Both tend to be influenced by interest rate changes. But the NME requires no theoretical postulate regarding the reasons for holding (or changing) money balances (e.g., speculative asset demand vs. transaction demand), or theoretical relationship to other hypothetical supply/demand schedules in the income determination analysis. Today the MDR and NME are probably affected far more by changes in government banking regulations and "compensatory" bank deposit requirements than by changes in financial speculators' liquidity preferences.

The Money/Income Multiplier (MIM) [ 5 ]
(back to contents)

The injection of new money (or NME) into the circular flow of income and spending is the primary exogenous element in income determination, and thus the effective element in the income multiplier. This is true no matter what the nature of the associated credit-financed spending -- whether it is for investment or consumption, or by business, consumers or government.

In the multiplier process, when borrowers spend the newly created money the recipients (on average) add to their money inventory (transaction balances) in proportion to the increase in their income, and spend the rest, or (by investing in credit instruments) lend it to others who (under present conditions) immediately spend it. This process continues until the total "leakage" from the circular flow into the public's additional transaction balances is equal to the initial injection of new money. In the process, the new money increases GNP by the reciprocal of the MDR -- currently, by nearly 7 times the amount of money increase -- much the way an injection of new bank reserves increases total checkable deposits by the reciprocal of the reserve ratio. Numerically, the MIM is equal to the traditional "velocity" (GNP/M1). However, as noted above, it has very different analytical significance.

This simple process is all there is to the "transmission mechanism" inside the traditional monetarist "black box." But this analysis obviously differs sharply from the monetarist assumption that "any first-round effects [of money growth on GNP] tend to be dissipated. [ 6 ]

The pervasive influence of the Keynesian model is evidenced by the fact that some of the most famous multi-equation computer models did not originally (and may still not) treat money growth as an independent (exogenous) variable.

The IDMF equation relating money growth to GNP growth
(back to contents)

gGNP   =  gM   -  gMDR

where "g" means growth rate

The GNP growth rate equals the excess of the money supply growth rate over the money demand growth rate. This is true whether GNP and M1 are both expressed in current dollar values or constant dollar values.

This dynamic relationship is illustrated in Figure 6, which shows how each of these factors affected the GNP growth rate (and resulting inflation rate) between 1960 and 1982.

Like the traditional equation of exchange ( MV = PT ) and the Keynesian NI equations, this equation is a conceptual identity, or tautology, in an ex post accounting sense. But a ante, in targeting GNP growth objectives, the Fed has to monitor (explicitly or implicitly) the adjusted trend value of the MDR and manage M1 accordingly. [ 7 ] (The Fed chairman, in his semi-annual reports, explains this to Congress almost that explicitly, though often in more veiled terms.)

It really doesn't matter very much whether the MDR (or "velocity") is basically stable (as the monetarists contend) or unstable (as the Keynesians contend), or what are the causes of its changes, as long as it can be, and actually is, monitored effectively on a current basis. Current monitoring of the MDR is particularly important when the factors determining it are undergoing significant change (as during 1980-89), whether these changes are short-run or long-run, and regardless of their causes.

Forecasting capability

To the extent that the main target of Fed monetary policy is the GNP growth rate, the IDMF analytical model is also useful for Wall Street Fed-watchers. The current trend of GNP growth can usually be predicted quite accurately (even several weeks before the end of the quarter) by monitoring the adjusted trend growth rate of the MDR and subtracting this from the adjusted M1 growth rate. (See Figure 7 [not yet Web-rendered].)

Interest rates
(back to contents)

Interest rates are the prices of credit, not "the price of money''. [ 8 ] They are determined, like other competitive prices, by the balance between supply and demand, both of which are today largely international. The IDMF analytical model requires no particular conceptual statement about the sources or reasons for either supply or demand.

However, with our credit money system, as noted above, about 10% (on average) of the total supply of credit is provided by new money. Thus, faster money growth tends to reduce interest rates directly in the short run, by increasing the supply of credit. But in a "cyclical" economic environment faster money growth may later tend to increase interest rates indirectly through the resulting faster rate of economic growth, since a faster rate of economic growth tends to increase the demand for credit more than it increases supply. (This tendency will be even more pronounced if the economic growth is so fast, as it approaches full employment, that it increases the inflation rate.)

However, in addition to these basic functional tendencies, actual market interest rates may also be affected by speculative expectations regarding future monetary policy, whether these are well-founded or erroneous.

The IDMF structural model
(back to contents)

The IDMF analytical model focuses major attention on the fact that most key individual economic flows and stocks have structural norms -- i.e. "normal" ratios to GNP, potential GNP or other relevant macro variables -- which tend to remain relatively stable (or change very slowly) over long periods of time. Figure 8 is a structural model of the estimated balanced-growth GNP ratios of the U.S. economy's money and credit, saving and investment flows. [ 9 ] A key research focus of this model is on efforts to determine the actual value of these apparent structural norms, and the functional reasons for these values and for their apparent stability and long-run trends.

However, in every economy, particularly one like ours which is not systematically managed, the actual GNP ratios of many economic factors tend to have relatively large short-run deviations around their apparent structural norms. In IEA POCKET CHARTS, deviations above and below the norms are shaded to facilitate their analysis.

Clearly, another of the main research focuses suggested by this model is to analyze these deviations and their functional causes, so as to devise current policies which will both compensate for the deviations in the short run (by offsetting deviations elsewhere) and also facilitate a return to normal values as soon as appropriate.

Elastic credit demand
(back to contents)

Until the "business cycle" is ironed out, each recovery from a recession causes deviations from the structural norms in consumer credit, inventory financing, etc., which provide abnormally large and unsustainable outlets for financial saving. This "elastic" credit demand inherently tends to decline or disappear entirely when the GNP growth rate returns to its normal full-employment trend value. To prevent recessions the inevitable decline in these "stimulating" elastic credit demands must be systematically offset by timely credit-balancing economic policy.

There are also longer-run structural-change elastic credit demands, such as caused by demographic shifts, changes in terms of credit (from 3-year to 4-year auto loans, introduction of credit cards and home equity loans, etc.), deficit-financed military buildups, construction booms etc., which also need to be offset in a systematic and timely manner by appropriate credit-balancing policy measures.

IDMF vs NBER "business cycle" analysis
(back to contents)

In an economy like ours with a steadily growing trend of potential GNP a growth-trend perspective is more functionally appropriate for analyzing "business cycle" fluctuations than the absolute standard of reference used by the National Bureau of Economic Research (NBER) and the Commerce Department's Bureau of Economic Analysis (BEA). See the article on "Economic Fluctuations in Growth-Trend Perspective for the IDMF conceptual framework for the definition and dating of recessions, recoveries, depressions and stagnation, and thus for measuring the relative leads and lags of individual economic series. As noted earlier, among the NBER/BEA "leading indicators" only money growth and housing have a genuine functional lead.

Institutional assumptions
(back to contents)

The IDMF model recognizes that all modem industrial economies are essentially managed economies (whether or not they are systematically managed). Take, for instance, the 1980-82 recession, the relatively stable recovery since then, and the concurrent U.S. federal deficit (which may have destabilized the world economy almost as much as the OPEC oil tax). All these were caused primarily by specific identifiable government policy actions (or deliberate non-actions), more than by inherent tendencies of an essentially autonomous economy. In fact, this has been true for every recession and recovery for at least the past 50 years. Thus, the key policy question today is not whether the economy is to be managed, but how -- i.e., on the basis of what empirical information, for what purposes, and with what political constraints.

The IDMF model is holistic in that it tries to comprehend all aspects of the real-world economy technical, institutional and political. It recognizes that an individual-enterprise, "free market" economy has the potential for incredible creativity, flexibility and adaptability -- including, for instance, accomplishing the necessary conversion from military to non-military production. But the model also recognizes that this adaptability can be greatly facilitated by credible government policy tools for maintaining the stable, non-inflationary, full employment overall economic environment necessary for making long-run business and personal plans and decisions.

"Supply-side" economics fails to recognize that capable businessmen can deal effectively with employment taxes, environmental regulations and other government interventions to the extent that the "rules of the game" are clear and consistent, that changes are gradual enough to facilitate adjustment within their planning horizon, and that the burdens of costly rule changes are fairly shared. Moreover, supply-side economics tends to ignore the fact that inflationary bottlenecks of productive capacity are largely due to the delayed effect of investment shortages during planned anti-inflation recessions -- and to the inability of businessmen to plan far enough ahead (compared to Jspan, for instance) because of high interest rates and chronic unpredictability of aggregate demand.


  1. Because of conceptual flaws, the FoF accounts do not yet provide the most appropriate data. Apparently even the Fed makes little analytical use of them. The poor data discourages more widespread use, and low demand for the data provides little incentive for research on conceptual and empirical improvement of the accounts -- a vicious circle. But the FoF accounts are potentially one of our most powerful analytical tools. Fortunately, quite useful estimates can be derived even from the present accounts.
    (back to ref 1)

  2. The financial deregulation which permits both interest payments on checkable deposits and limited third-party checking on money market accounts, has fuzzed the empirical separation of money and credit. This separation can probably be reestablished now only by requiring 100% reserves on checkable deposits. But for most policy purposes the problem can be minimized by close monitoring of the Money Demand Ratio.
    (back to ref 2)

  3. One reason the Fed's new r "inflation indicator" is inappropriate as a basis for counter-inflation monetary policy is that it uses M2 as its definition of "money." The three fourths of M2 which are credit assets can't be directly controlled by the Fed and are not a functional cause of inflation.
    (back to ref 3)

  4. The IDMF money-inventory concept differs from the Cambridge cash-balance approach in that it views money stock mainly as a necessary "cost of operations" rather than as a discretionary form of wealth. With our present highly developed financial markets, where high interest rates can be obtained on even immediately available credit assets, it probably applies equally to Keynes' individual and institutional speculators.
    (back to ref 4)

  5. This money/income multiplier should not be confused with the checkable-deposit reserve multiplier (the reciprocal of the reserve ratio) which, is sometimes referred to as the money multiplier.
    (back to ref 5)

  6. Milton Friedman, New Palgrave Dictionary of Economics, v III, p.10
    (back to ref 6)

  7. Unfortunately, with our present credit-related system of money creation and irresponsible fiscal policy, functionally appropriate GNP targeting is in practice constrained by its side effects on interest rates and the dollar exchange rate, and by the inadequacy of our non-monetary anti-inflation tools.
    (back to ref 7)

  8. Credit, of course, is the rental of money. But the term price usually refers to the terms of sale (purchase), not rental.
    (back to ref 8)

  9. This is a comprehensive model, derived from 1947-63 data. Each component norm is estimated in such a way that each aggregate norm is equal to the sum of the component norms. This model clearly needs to be updated. (For example, data for the now-crucial Rest of World sector were then too small to be significant and were, in effect, netted out.) But later estimates for individual series in IEA POCKET CHARTS suggest that many of the values tend to remain surprisingly stable over long periods.
    (back to ref 9)

Written: April 10, 1990
Posted: July 12, 1998 (with slight revisions)
To IEA home page