Reserve Bank’s Analysis of Money Supply!
The RBI does not follow explicitly any theory of money supply either in its verbal explanations or in its data presentation.
Instead, it publishes every month a purely accounting analysis of what it calls factors affecting money supply, or ‘sources of change in money supply’ in the form of a table in its Bulletin.
All the so-called money-supply analysis of the RBI, other official agencies, and unfortunately also of most academic economists follows uncritically this table. Thus, a purely accounting analysis has usurped the place of any theory of money supply. This is most unfortunate for a proper understanding of the forces operating on money as well as for the correct formulation of any monetary policy. Before offering a fuller explanation of these points, let us see what the RBI’s accounting analysis I’s. This is reproduced below.
Sources of Change in Money Supply:
1. Net hank credit to government (A + B)
A. RBI’s net credit to Government (i-ii)
(i) Claims on government
(ii) Government deposits with RBI.
B. Other banks’ credit to government.
2. Bank credit to commercial sector (A + B)
A. RBI’s credit to commercial sector
B. Other bank’s credit to commercial sector
3. Net foreign exchange assets of banking sector (A + B)
A. RBI’s net foreign exchange assets
B. Other bank’s net foreign exchange assets
4. Government’s currency liabilities to the public
5. Net- non-monetary liabilities of banking sector (A + B + C)
A. Time deposits with banks
B. Net non-monetary liabilities of RBI
C. Other net non-monetary liabilities of banks.
In terms of the above factors, then,
M = 1+ 2+ 3+ 4-5.
The above table is derived by taking the consolidated balance, sheet of the banking sector as a whole (the RBI and banks) and breaking down its financial assets sector-wise in the same manner as the ‘factor affecting H’.
The main points of criticism of the aforesaid (accounting) analysis of money supply are summed up below:
1. Since the RBI’s analysis is a purely accounting or ex posts one, it has no explanatory power of its own. That is, it cannot ‘explain’ the consequences (for money supply) of various policy and non-policy autonomous changes, such as the open market operations of the RBI, or changes in the statutory reserve requirement for banks, or changes in net foreign aid, etc., much less explain how and why such autonomous changes affect the supply of money.
All that it does is measure changes in the stock of money after they have occurred (not predict them beforehand) and allocate the measured change to different sectors, factors, or sources according to a particular scheme of classification. But a mere classification of the data provides no explanation.
Nor is a particular accounting scheme of data presentation any substitute for theory. Nor does it obviate the need for a theory, however crude, for understanding the money supply mechanism or the way several forces operate to bring about changes in the supply of money, and for monetary planning. Obviously, before the RBI uses a particular monetary-control measure, it must know how this measure is likely to function and its likely effect on the quantity of money supply.
We may support the above argument by an analogy drawn from the well-known fields of national income measurement and determination. It is well accepted by now that national income measurement is a problem of (social) accounting and that national income determination is a problem of economic theory. Since ex post total income is identically equal to total expenditure in a closed economy, one; way to measure national income (Y) is to measure total national expenditure.
Among several possibilities, the latter can be taken as the sum of private consumption expenditure (C), private investment expenditure (I), and government expenditure (G). This gives Y= C + l + G. Thus, income is measured in terms of the components of expenditure. From the income-expenditure identity if follows arithmetically that, exposit, increase in any one of the component expenditures, C, I, or G, will be matched by an equal increase in Y.
But this is a purely tautological statement, true by definition, and thus devoid of any explanatory content. It does not tell us whether ex ante it is the increase in C which leads to an increase in Y, or vice versa, or it is the increase in some other factor which lies at the root of increase in both C and Y.
But this is what a genuine theory should explain. For example, the Keynesian theory of national income determination explains what factors lead to changes in Y and why. In its simplest form, both I and G are assumed to be autonomous of Y. Together they are called autonomous expenditure (A). C is assumed to be an increasing function of Y.
Since A is assumed to be autonomous of Y, changes in A can occur independently of changes in Y. It is these changes in A which bring about induced changes in Y and C via the well-know multiplier process. Thus, changes in both C and Y are explained by autonomous changes in A.
These arguments are fully applicable to the analysis of factors affecting money supply. We have stated earlier that the RBI analysis is only an exercise in ex post measurement and therefore cannot serve the purpose of a theory. But if we are genuinely interested in explaining changes in money supply, we must have recourse to some theory of money supply.
In the first instance, such a theory attempts to explain changes in M in terms of changes in H and in m. In addition, it tells us what lies behind changes in m and how to explain them in terms of their determinants proximate and ultimate.
Changes in H were traced to their several sources, using only accounting analysis. In a fuller study, the nature of each source can (and should) be examined and explained either as a policy variable, as an endogenous variable, or as an exogenous variable.
2. The RBI’s accounting analysis adds up components of H and M. ignoring the dependence of the latter on the former. The result has been a total denial of any money-multiplier process or of the ‘secondary’ expansion of money and credit by banks induced by reserves accretion— something which any monetary economist or even the RBI has never seriously denied.
As a result, the accounting table has been a rich source of erroneous propositions, some of which are examined below:
(i) That ‘the government sector’s total impact on money supply is equal to its total budget deficit minus its net purchase of foreign exchange from the RBI’. This is wrong, because what the latter sum measures is the effect on H, not on M. The supply of M increases by a certain multiple of the increase in H, depending upon the value of the money multiplier m;
(ii) That ‘increase in other banks’ credit to the government leads to an equivalent increase in money supply’. This is not correct, because increase in other banks’ credit to the government only reallocates banks’ credit in favour of the government at the cost of the commercial sector, leaving the total supply of M as well as banks’ credit practically unchanged. The last part of this statement is explained under the next point;
(iii) That bank credit to the commercial sector is an important factor contributing to the expansion of money supply. Bank credit in such statements usually includes both the RBI credit and the credit from banks. The causal role of the former in the expansion of money supply is understandable; but not so the role of the credit from banks. The latter is as much a dependent variable as the total money supply.
Therefore, to say that money supply increased because banks’ credit (whether to the government or the commercial sector) increased is not true. Nor does it explain anything. It only moves the enquiry about the money-supply increase one step backward. Now, the question must be asked: how could banks increase their credit? Do they have unlimited power to increase credit or is it limited by something? To say that banks could increase their credit because their deposits grew clearly contradicts the earlier assertion that an increase in bank credit causes an increase in money supply (which includes demand deposits of banks).
Also, it is no solution to say so, because then we have only to re-word our previous question and ask: why did the bank deposits grow?? Either way we must locate some autonomous change which induces banks to increase their credit and, in the process, deposits as well. The answer is provided by the H theory of money supply.
3. The RBI’s accounting table suppresses completely the RBI credit to banks as a factor affecting money supply, because in consolidating the assets and liabilities of the banking system as a whole it gets cancelled out as an internal transaction. But this is patently wrong, because the borrowings of banks from the RBI do increase H and thereby M.
The only way open to the RBI to disabuse money-supply analysis in India from these and similar errors and consequent faults in policy-making is to jettison its accounting approach to money-supply analysis and its concrete embodiment in the accounting.