I find Jérôme de Boyer des Roches's latest contribution to this 
discussion quite unhelpful.  He writes:
>
> We have not to forget the price level.
>
> Indeed, in Keynesian economics we have four unknowns (Y, R, P, W) and 
> four equations (the first postulate which links the real wage rate 
> (w=M/P) with the level of income and the equilibrium conditions of the 
> market for goods (IS), for cash balance (LM) and for financial assets 
> (BB)), but only three independent equations (see de Boyer, JHET, vol. 
> 32, n°2, June 2010). Therefore, in order to define a Keynesian 
> equilibrium, we need an additional equation. Hicks (1937) introduced a 
> given nominal wage rate, Hansen (1953) a given level of prices, 
> Patinkin (1965, chap 10) a given level of income. I understand that 
> Robert Leeson resumes this last hypothesis.
>
How meaningful is it to be discussing "equilibrium" or an equilibrating 
process by assuming "a given nominal wage rate" (Hicks)?  Who gives that 
value?  Aren't wage rates determined in the various employment centers, 
as A.C. Pigou (1933) well cautioned economists to keep in mind?  Isn't 
the price level determined by the supply and demand for money 
(currency), as we learn from David Hume(1752) on down to Alfred Marshall 
(1923)?  And how meaningful is it to assume a given level of income 
(Patinkin), when the national income is the aggregate of incomes earned 
by individuals in the various workplaces?  To me, these exercises are 
all the result of the ascription to Keynes's economics that is far 
removed from the economics of the marketplace.

Jérôme continues with:
>
> Now, if the level of income is not variable, but the interest rate, 
> the price level and the money wage, what happens when the propensity 
> to save increases ?
>
> The multiplier is at work: consumption decreases, demand for goods 
> decreases, the IS curve shifts to the left and its slope increases. 
> Because there is excess supply of goods, price level decreases, demand 
> for money decreases and demand for financial assets increases (LM and 
> BB move downwards), interest rate decreases, investment (demand for 
> goods) increases (and supply of financial assets increases), etc…. 
> Again, the multiplier is at work until the supply of investment goods 
> (S) is equal to their demand (I). At the end, we have the given level 
> of income (unchanged), lower interest rate, higher levels of 
> investment and saving, lower level of consumption, lower levels of 
> prices and nominal wage (but the wage rate remains the same), and 
> higher value of cash balances. Note that the unchanged level of income 
> is not an outcome, but an hypothesis.
>
Savings constitute the "demand for financial assets," as Jérôme 
correctly notes in his JHET article, p. 268.  That is why Keynes's 
multiplier argument, founded on the notion that savings are a withdrawal 
from the expenditure stream, is not a meaningful proposition.  But then 
again, Jérôme also uncritically believes the Keynesian view when he 
writes (p. 267) that, "savings, represent[...] a leakage from the 
mechanism whereby demand tracks income."

When one recognizes that savings are spent, "and nearly at the same time 
too" (Smith, _WN_; see also Mill, _Works_, 2: 70), one would recognize 
that increased savings do not decrease "aggregate demand" (C + I + G). 
With a non-zero interest elasticity of demand for loanable funds, 
increased savings decreases interest rates and increases the quantity of 
loanable funds demanded to be spent.  (Of course, there are those, like 
Joerg Bibow, _HOPE_ 2000, who accept and defend Keynes's view that the 
loanable funds theory of interest is a "fallacy.")  Even if some 
borrowers pay down their debts (as Michael argues), the lenders now have 
the funds to lend to other people.  Only when one defines, incorrectly, 
as cash hoarding would increased saving shift the so-called IS curve to 
the left.  We safely can leave the unhelpful IS-LM model to the teaching 
of intermediate macroeconomics.

Now, Keynes himself separates savings from supplying investment funds or 
promoting increased investment spending.  He declares that the "supply 
[of 'liquid resources'] depends on the terms on which the banks are 
prepared to become more or less liquid ... saving does not come into the 
picture at all" and "The investment market can become congested through 
the shortage of cash.  It can never become congested through the 
shortage of saving.  This is the most fundamental of my conclusions 
within this field" (_EJ_, December 1937). These declarations are in 
response to D.H. Robertson's criticisms of Keynes's view of savings.  
I'm therefore puzzled by Jérôme's attempts to reconcile Keynes's 
economics with the discussion about savings supplying the funds for 
investment spending.

Regarding Roger's reference to Currie's work, I do not consider it 
relevant.  Currie was concerned about stabilizing the "means of payment" 
(M1) rather than assuring the flow of savings into investment spending.  
Indeed, it is such concern over the stability of M1 that led to the 
unfortunate recommendation of a 100% reserve requirement, endorsed also 
by the likes of Irving Fisher and Milton Friedman.  But with the 
implementation of a 100% reserve requirement, the portions of demand 
deposits that banks could otherwise lend to investors would be 
eliminated.  That is why the fractional-reserve banking is more 
favorable to promoting investment (not only the purchase of capital 
goods, as Keynes's definition assumes and other discussants here keep 
using) than the recommendation to which Currie's argument leads.

James Ahiakpor
-- 
James C.W. Ahiakpor, Ph.D.
Professor
Department of Economics
California State University, East Bay
Hayward, CA 94542
(510) 885-3137 Work
(510) 885-4796 Fax (Not Private)