§6 A Flow-Chart to Disprove the Keynesian Multiplier Effect Theory

Keynes' principle of effective demand based on the the Keynesian multiplier effect is mathematically wrong .

The Keynesian multiplier effect does not exist.

Yuichiro Hayashi, Y. Hayashi

July 1, 2004  Publication of the first draft

"The General Theory of Employment, Interest and Money" was published in 1935-36 by John Maynard Keynes. From that time, the Keynesian multiplier effect theory has constituted one of the bedrocks of current economics. For several decades, the theory has been one of the main themes not only in textbooks but also in the real economy. Current national economic policies are largely affected by it. Every book and web-site relating to economy is, except for a few objections, described based on the principle of effective demand which is connected to the investment multiplier effect( Fig.3-3 ) .

The objections against Keynes's multiplier effect theory are either arguments without mathematical disproof* or mentions that the investment multiplier 1/ (1 - MPC) exists but  its effect is small.

* On 11 July, 2009, the author suddenly received an e-mail from Dr. Brian Chapman in Australia. According to the e-mail, he also has researched errors on Keynes' multiplier effect theory for long years, and he found my website on the evening before. Refer to the following websites. His conclusion that Keynes' multiplier effect formula is wrong is the same as mine. It was verified mathematically and independently of my research.

See the equations in [Question-18]. The basic equations of the present discussion are written only on 2 lines:Y2 = C2 + I2 , ΔY2 = ΔC2 + ΔI2, ΔC2 = ΔC1 = a K ΔY1, ΔY2 =ΔY1, ΔY1 - a K ΔY1 = ΔI2, ΔY1 = ΔI / (1 - a K). If a K =0.6, we have ΔY 1= 2.5ΔI2.

Fig.3-3 looks beautiful. Both the above equation and the one in which ΔI is replaced by ΔI2+ ΔNX2 + ΔG2 in place of ΔI2, are surely identities. Any error is nowhere to be found. Everybody has acknowledged scholastically (mathematically) the truth of the multiplier effect theory for decades. Even if anyone doubts the effect of the Keynesian multiplier in a depression, nobody can disprove it mathematically and presents a new theory displacing it.

Although this author recognize the existence of Keynes's investment multiplier as an equation, he insists that it dose not mean the the multiplier effect expected by Keynes, i.e. that the investment multiplier effect does not exist mathematically. Even I doubted the truth of my argument at first. I shall  present my questions here and answer them.

The Keynesian multiplier effect chart is shown in Fig.3-3.

Fig.3-3 Keynesian multiplier effect chart

(1)  On the one hand, the notation ΔY2 is used for showing the sum of the incremental components in GDP in the reference year as shown in Eq.(3-37). On the other hand,  the notation ΔY1 is also used in order to show the equivalent value to, on Y1 coordinate, for example, only ΔG2. Unless otherwise stated, the former definition is adopted:

 Δ Y2 = Δ C 2 + Δ I2 + Δ NX2 + Δ G 2 (3-37)

,where, for simplicity of notations, Δ government consumption = 0 is assumed.

(2)  Readers should know that the partial 'gross value added (GVA)' is a person's economic activity value which is added to valueless resources. Corporation is included in the word person.  All economic activities including producing, earning income and expending, are expressed by the economic those of partial GVA. As is well known, the GVA is measured from various economic aspects. For example, ΔC2 (consumption goods) is a part of ΔY2 which is measured in the economic aspect Y2 (final products). It has the name consumption (goods ), in which its partial GVA is collected in the name of consumption.

(3)  Readers should recognize that each component in the right side of Eq.3-37is a realized final product within a certain time period ( or that which will be realized within a period between now and a future day ). It should not be allowed, in the equation, that one product is already realized and another is not yet produced and will only be realized on a certain future day. If the readers intend to extend the period from the end of the present period to a future day, they should adjust all component data in the equation to that future data.

(4)  Readers should recognize that the equations ΔY1 (incomes) = ΔY2  (sales of final products) = ΔY3(all costs less purchase of intermediate products + profits) will hold in a yearly period in economic theories. Furthermore, when we have the equations, ΔY1=∑components1, ΔY2=∑components2 and ΔY3=∑components3,  the readers should recognize that, for example, the equation ∑components1 = ∑components2 holds, and we must not subtract any components from only one side of this equation.

(5)  Readers have recognized economic characteristics of ΔY1, ΔY2 and ΔY3. Each indicator has its components( parts) respectively. For example, ΔY3 has compensation of employees, operating surplus etc,  and ΔY2 has final products of secter-1, secter-2 - - -  industries. The readers should recognize that each component( e.g. ΔC2 ) in one aspect ( e.g. ΔY2 ) can be conceptually  expressed with a composition of another aspect's components ( e.g. compensation of employees and operating surplus, - - -etc.), if a theoretical means is established. This is the concept adopted in the input-output table. In Chapter2, the new matrix equation, [vector of ΔY2 (GDP) components] =[transformation matrix ] · [vector of ΔY3 (GVA) components] is presented by the author. Consequently, the multiplier effect concept( the second impact effect by an incremental income) is not essentially included in the input-output table.

(6)  Final products, incomes and expenditures are simultaneously realized within one year. If a government's additional expenditure ΔG1( = ΔG2 ) is executed, the money for ΔG1 has been obtained from the incomes within the one year, via transfer or borrowing. Note that recipients of ΔG3 ( costs, such as employees compensation) do not necessarily correspond to tax payers or fund providers.

(7)  As far as we allow to adopt the Assumption 5 (Δ I 2+Δ NX2=0 or Δ I 2  = Δ NX2 = 0 ) shown in §3, readers will surely recognize that the equation ΔY1 = ΔG2 +ΔC2 holds at the Keynesian cross ( this is really at any time). The readers know the solution of the income ΔY1 = ΔG2  / (1 - aK) known as the Keynesian multiplier effect. The equations ΔG2  / (1 - aK) = ( ΔY1 - ΔC2)  / (1 - ΔC2  / ΔY2 ) = ΔY2 = ΔG2 +ΔC2 hold.

The Keynesian multiplier effect equation is understood in such a manner that the multiplied income ΔY1 = ΔG2  / (1 - aK) will be obtained  some day in the future naturally or in an induced manner, by the execution of ΔG2. The above-mentioned equations show, however, that after all the government should  realize ΔY2 = ΔG2 + ΔC2, which is a product or the breakdown of the product, in order to realize income ΔY1 = ΔG2  / (1 - aK) within an accounting period in which ΔY1 and ΔY2 are measured.

To tell the truth, we have overlooked the fact that ΔC2 is already generated simultaneously with the execution of ΔG2 in Fig.3-3  over a period of years (*).  Fig.3-3 must be exactly expressed as shown in Fig 4-6 by putting ΔC2 and correcting ΔY2 = ΔG2 into ΔY2 = ΔG2 + ΔC2 in Fig.3-3.

[* It will be shown later that this results from the equilibrium theory method.

Fig 4-6 Corrected Keynesian multiplier effect chart

However, it is not the government's ( or Keynesians') purpose to create, naturally or in an induced manner, ΔY1 (= ΔG2 +ΔC2) from only ΔG2. Consequently, Fig.4-6 is mathematically correct, under Assumption 5, but it is also  incorrect/ flawed for only expenditure ΔG2. In addition, even if we allow that Fig 4-6 shows the correct relation for both ΔG2 and ΔC2, we can't, by any means, include ΔC2 = 0. Thus, Fig.3-3 is doubly incorrect/ flawed.

(8)  Someone may say, "We know that we can't create income ΔY1 = ΔG2 +ΔC2 from only expenditure ΔG2 in a short time. We expect that ΔC2 will be obtained some day in the future according to the 'propensity to consume' or it will be induced by consumer confidence". This claim is wrong( incorrect/ flawed). The consumption ΔC2 which is expected to be induced in the future is already finished as it is written in national economic accounts in the year when ΔG2 is executed.  In one year, ΔG2 had had an effect on the economy. Very much so!

The relationship between output (final products ) and the corresponding input( GVA) is shown in Fig.3-14. In a business depression, private investment will decrease.  ΔG2 due to borrowing might have been a partial firms' profit or a partial employees compensation, both of which should have decreased, if ΔG2 due to borrowing had not been executed.

Fig 3-14 Input-Output Table chart

(9)  Readers should know that the ΔG2-Y1 problem shown in Fig.3-3 is slightly different from the problem shown in [Question-18] in §4. I could not strictly discriminate between them at first. In the former case,  ΔI2 = ΔNX2 = 0 is assumed, but in the latter case,  ΔG2 = ΔNX2 = 0 or, ΔI2 including ΔG2 and ΔNX2, is assumed.

(10)  Have readers understood the explanation mentioned above? In fact, this explanation also has a large error. In short, both  Fig.3-3 and Fig.4-6 are absolutely wrong (incorrect/ flawed ) regardless of the reason that we can't include ΔC2 =0 in these figures. These two Keynesian figures were obtained by adopting Assumption 5 ( Δ I2 = Δ NX2 = 0). Therefore, we obtained ΔY1 = ΔG2 / (1 - aK). In fact, we can't adopt Assumption 5. The reason is that ΔY1 = ΔG2  / (1 - aK - aI - aNX), where aI = 'marginal propensity for private firms to invest', should hold. If we take aI into consideration, the Keynesian multiplier will become about 10 times( see [Answer-24].).

This will be explained in a little more detail. We will consider two cases: Case-1······ the analysis under the condition that Assumption 5 ( Δ I2 = Δ NX2 = 0) does not exist; Case-2······ the analysis under the condition that Assumption 5 ( Δ I2 = Δ NX2 = 0) is adopted. The marginal propensity to consume is defined as aK = ΔC2 / ΔY2. If we suppose aK = 0.6, the equation (Δ I2 +Δ G2+Δ NX2 ) /ΔY2 = 0.4 holds with Case-1.

With Case-2, ΔG2/ΔY2 = 0.4 must hold when aK = 0.6. This is impossible in the real economy. The real amount of ΔG2 /ΔY2 will be close to a statistical amount of G2 /Y2. If we suppose ΔG2/ΔY2 = 0.1, then, the chart, which includes both ΔG2 and ΔY2, will have to be drawn in such a manner that the amount of ΔG2 is about 1/10 of ΔY2, i.e. ΔY2 =10ΔG2.

See Fig.4-6. The ratio of ΔY2 to ΔG2 is drawn as 1 to 0.4. This resulted from neglecting the existence of both Δ I2 and Δ NX2. Fig.4-6 is absolutely wrong( incorrect/ flawed). This has occurred when the amount of Δ I2 +Δ NX2 has been taken away from only the right side of Eq.(3-37) (the amount Δ C2 + ΔI2 is a part of the right-hand side of Eq. (3-37)), although the definition of ΔY1 doesn't change ( ΔY1(=ΔY2) is the whole of the right-hand side of Eq. (3-37)). This has brought  ΔY2 ≠ΔY1 which doesn't satisfy the Keynesian cross condition. Therefore, an error in the Keynesian multiplier effect theory has originated in the adoption of the assumption, Δ I2 = Δ NX2 = 0. [ This sentence on 3 lines might be rewritten. ]

We find that such a discrepancy between the right-hand side and the left hand side was not caused, if we had defined Y1 as Y1 =C2 + Gfrom the beginning. However, we might make an unrealistic theory with the assumption that private investment doesn't exist. Consider a system which consists of components xi ( i =1,2,·  · ). The whole is included as one of components. Suppose that the change of a certain component xk is not effected by small changes of the other components. For example, this is the case such that only the investment is a function of an interest rate. When we make a differential analysis to the components not relating to the interest rate, we shall easily understand that both the investment itself and this amount in the whole must be taken away at the same time.

Furthermore, it is known today, from the break-even analysis for an income statement, that we have the two types of wages: One is proportionally variable wages to the quantity of the production; The other is fixed type wages to it. Consider the case where one wishes to analyze a problem which consists of a production( = sales), costs with wages and a profit.  If we make a differential analysis to this problem without special attention, the fixed type wages might be taken away from the analysis. These are the dreadful aspects in differential analyses. We must review whether these treatments might have been done in other theoretical, economic analyses. [This paragraph would be reconsidered later.]

(11)  Let a( t) be a nonlinear function of time t. We define notations as: aK( t)= ΔC2( t)  / ΔY1( t)  aI( t) = Δ I 2( t) / Δ Y1( t),  aNX( t) = ΔNX2( t) / ΔY1( t) and  aG( t) = ΔG 2( t) / ΔY1( t), where aK( t) or aI( t) etc. is the marginal propensity of expenditure to each good. Since aK( t)  =  ΔC2 ( t) / Δt  / (ΔY1( t)/ Δt) at any t, we have

 aK( t)  +  aI( t)   +  aNX( t)  +  aG( t)  =1 (3-R7)

Multiplying both sides of Eq. (3-R7) by ΔY( t) ( = ΔY1( t) = ΔY2( t) ) again gives

 ΔY( t) = aK( t)ΔY( t)  +  aI( t) ΔY( t)  +  aNX( t)ΔY( t)  +  aG( t)ΔY( t) (3-R7-1)

Thus we obtain, for example, from Eq. (3-R7-1), the following equations:

 ΔY( t) = (Δ I ( t)  +ΔNX( t)  + ΔG( t) ) / (1 - aK( t) ) (3-R7-2)
 ΔY( t) =  ( ΔC ( t) +Δ I ( t) ) / (1 - aNX( t) - aK( t) ) (3-R7-3)

Eq.  (3-R7-2) expresses the following two equations:

 (Δ I ( t)  + Δ NX( t)  + ΔG( t) ) : ΔY( t) = (1 - aK( t) ) : 1 (3-R7-4)
 ΔC ( t) : ΔY( t) = aK( t) : 1 (3-R7-5)

That is to say, both Eq.(3-R7-4) and Eq.(3-R7-5) as well as Eq.(3-R7) express the proportions of the parts ΔC ( t), Δ I ( t), Δ NX( t) and ΔG( t) to the whole ΔY( t).  For that reason, Eq. (3-R7-2) is an identity.

In addition, assume that indirect taxes = 0 to simplify notations. ΔGVA is shown as: ΔGVA = ΔD + Δπ + ΔW  + ΔT - ΔT = Δπ + ΔW where D is depreciation, π is operating surplus, W is employees compensation and T is transfers including direct taxes. From the relationship between GVA and GDP, we have, omitting the symbol ( t),:

 ΔY =  (Δπ + ΔW ) + (ΔC + ΔI +  ΔNX + ΔG) - (Δπ + ΔW)+ (ΔT - ΔT)    =  (Δπ+ ΔW ) + {ΔI + (ΔG - ΔT) +ΔNX} - (Δπ +ΔW - ΔT - ΔC) (3-R7-6)

As mentioned in (10), why can't we adopt the assumption, Δ I2 = ΔNX2 = 0 ? The reason is that Eq.(3-37) should hold between the incremental five goods, ΔY2, ΔC2, Δ I 2, ΔG2 and ΔNX2. This is the same as Eq.(3-R7) holds. This means that each good should be generated as follows:

(a)　It is generated in such a manner that Eq.(3-37) or Eq.(3-R7) is satisfied;

(b)  It is by nature independent of each other. However it is generated in a half mixed way of both dependence and independence, observing with each other and being affected by each other;

(c) In Eq.(3-R7-6), it is generated satisfying the following constraint condition:

 ΔI + (ΔG - ΔT) +ΔNX = Δπ +ΔW - ΔT - ΔC (3-R7-7);
(d)  It is not necessarily satisfied that each term in Eq. (3-R7) equals always 0;

(e)  The relationship between each economic term in the equations concerning the three economic aspects is realized in such a manner that it satisfies the break-even chart described in (19) below, reflecting people's will.

Consequently, the following economic phenomena will often occur in theory: for example, no amount of an increase ΔG2 can give an increase ΔY2; a little amount of increase ΔG2 can generate the other goods largely.This is the basic principle in human activities.

(12) If we adopt the next assumption:

Assumption7: ΔG2 = ΔNX2 = 0

, we obtain a figure in which ΔI2 is replaced in place of ΔG2 in Fig 4-6. In this case, the very problem of the additional expenditure by a government dose not hold because ΔG2 = 0.

(13)  If we adopt the following assumption:

Assumption8: ΔG2 +ΔNX2 is included in ΔI

,we obtain ΔY2 = ΔC2 + ΔIT. This case is like the one taken up in (12). The figure for this economic model is gotten by replacing ΔIT in place of ΔG2 in Fig 4-6, which is shown in Fig.5-5 below.

Fig 5-5 Keynes’ original chart

This economic model is used in J. M. Keynes' original work, although he didn't show its model chart). In this regard, he thinks that Δ2 is naturally provided proportionately to ΔIT someday without determining the time period.  We shall name Fig 5-5  'Keynes' original chart'.

In Keynes' original chart, the Keynesian cross condition is satisfied and so the MPC exists. Does  any strange result occur when we use the chart to analyze the actual economic phenomena?

To conclude ( see 5.3 and 5.4 in the next section), we can't use a chart like Keynes' original chart in order to analyze economies over time if time changes. Keynes' original chart model is a two-degrees of freedom system in which the variables are both C2 and I2, and Y2 is a dependent variable. Just when Keynes assumed the existence of the MPC, one constraint condition is added to the system to create a one-degree of freedom system.

J.M.Keynes claims the multiplier effect ΔY1ΔIT/ (1 - MPC). Then, ΔIT must be the  only remaining variable in the one-degree of freedom system. ΔIT should be given as a variable which doesn't relate to the passage of time at the beginning time point before the execution of ΔY2. ΔC2 depends on ΔIT, and so both ΔC2 and ΔIT  don't relate to the passage of time. Thus, the economy shown by Keynes' original chart doesn't relate to the passage of time . This relationship is explained in detail in Fig.5-4 in the previous section. By the way, we can select ΔC2 as a remaining variable, when we obtain ΔY1=ΔC2 / MPC. Analogously, ΔY1 depends on ΔC2, and both of them don't relate to the passage of time.

Fig 5-4 One-degree of freedom system

By the way, we can select ΔC2 as the remaining variable, when we obtain ΔY1=ΔC2/ MPC. Analogously, ΔY1 depends on ΔC2, and both of them don't relate to the passage of time. We have a system which consists of ΔY1(=ΔY2), ΔC2 and ΔIT , That the system is a one-degree of freedom system is as follows: the actions of the system is determined uniquely only by the action of the whole ΔY2, and ΔC2 and ΔIT are the constant proportions to the whole ΔY2, even though each variable is a function of time.

(14)  In Keynes' original chart theory, the whole ΔY2 , i.e. both ΔC2 and ΔIT doesn't change with time but rather by the uncontrollable fluctuations of people's decisions. Furthermore, ΔIT must always be provided before the finished change ΔY2. In short, the economy expressed by Keynes' original chart is one where both IT  and C2 don't relate to the passage of time. This society is one where ergodecity is assumed. This might be similar to a preindustrial society.

This economic model might be appropriate in analyzing a distribution problem of economic goods or incomes, but it is not fitted  for analyzing a dynamic economic problem of an unsteady economic situation in a capitalist society. This viewpoint is important in particular when analyzing big problems of the capitalist country in an unsteady state of economy, most of all, an unemployment problem.

(15)  A large shortcoming in Keynes' original chart theory is that the MPC condition is broken just when we intend to supply ΔIT . To resolve this problem, we must simultaneously and instantly provide ΔC2 (=ΔIT∙ MPC /(1-MPC) according to the supply ΔIT which doesn't relate to the passage of time.

(16)  The next large shortcoming in Keynes' original chart theory occurs just when ΔG2 is supplied in order to resolve an unemployment problem. This will be explained using  Fig.5-6. Let ΔIT be ΔIT = ΔI2 + ΔNX2 + ΔG2. When we select ΔG2 as a variable corresponding to the one-degree of freedom in this analysis, we have to consider the roles of remaining ΔI2 and ΔNX2 at the same time, and have to take them into consideration in the analysis.

Fig 5-6

By virtue of Keynes' original chart theory, the equation  ΔY1= (ΔI2+ ΔNX2 + ΔG2) / (1 - MPC) can be derived. The MPC is defined for ΔIT. Consequently, both ΔI2 and ΔNX2 as well as ΔC2 have to be simultaneously and instantly provided with the production ΔG2. This brings up a puzzling situation. Now that we can't produce ΔI2 and ΔNX2 instantly (private investment demand decreases), according to the supply ΔG2, the multiplier effect theory has been used for analyzing an unemployment problem.

In Keynes' original chart theory, simultaneous and instant supplying of ΔI2, ΔNX, and ΔG2 is sought according to the supply ΔG2.

Can readers claim that Keynes’s original chart theory is an effective ways to analyze the problems of a dynamic economy though this theory has such characteristics? If we can't, Keynes’s original chart theory will lose its effectiveness at least in regard to a dynamic economic problem analysis.

(17)  As mentioned above, the real economy will have many-degrees of freedom with respect to time.  However, the MPC is factually found by statistical observations of the real economy as shown in Fig. 3-11, in which ak = MPC . By the way, it is said that we have two types in consumption lines. One is the Keynes type in which the consumption line is a linear one with a vertical axis-intercept. The other is the Kuznets type in which the line is linear through the origin. The line  will really be a nonlinear line through the origin. How can the MPC be observed in the real economy?

Fig 3-11 Corrected Keynesian multiplier effect model

The author thinks that the MPC can be obtained only when we look at  the resultant moving of consumption in a many-degrees of freedom economic system in a stable economy. Each firm always endeavors to stay profitable. The MPC, derived from the consumption psychology of a community exists. Each consumer observes his present income and envisages his future life. Each entrepreneur observes the consumption behavior of people and envisages his future state. Everyone observes present business conditions and decides his daily behavior.

The moves of a consumption product and an investment product are similar to those of a right leg and a left leg. If a road ahead is bright and seems to be safe, their moves will appear to be constant. The walking in this case will look like as if both legs are one pair and the move is based on a one-degree of freedom system. However, if one takes a right step on a road ahead being dark and not safe, the left step may stop or back off after that. If one is in the middle of boxing, each leg will move freely and independently. In the latter two examples, both legs move in a two-degrees of freedom system.

In a similar way, a consumption product and an investment product will move in the stable economy in such a manner that the MPC is kept constant. However, in an unsteady economy with respect to time, each product will probably move independently by free-will, and the original nature of products, i.e. independence will appear.

(18) Just as the Keynesian multiplier 1/(1- MPC) doesn't exist, the tax multiplier MPC/ (1 - MPC) doesn't exist also. This stands to reason because taxes are partial transfers between income group. See the explanation on the taxes problem in 5.1.

(19) In fact, there exists a more fundamental defect in the Keynesian multiplier effect theory, although this defect is not a mathematical error. As described in the previous section, there are two types of charts when we analyze economic problems. One is a time-series( or proportion) type input-output chart shown in Fig.5-1, and the other is a break-even type chart shown in Fig. 5-2.

Fig 5-1 Time-series type input-output chart

In the time-series type chart shown in Fig.5-1, the horizontal axis shows realized, cumulative production output. The vertical axis shows components of realized, cumulative GVA( input). The time-series amount can be a monthly cumulative one, or a yearly final one to show it over a period of years.

The characteristics of the time-series type chart are as follows:

(1) The band width of operating surplus( 'profit' in Fig.5-1) shows a part proportion of the whole, Y1. This holds for other components.

(2) As the operating surplus is a function of one variable, Y2, Y2  is uniquely determined for a given operating surplus, if the economy is in a stable condition.

(3)  Incremental GVA, ΔY is expressed as follows: ΔY = Δπ +ΔW +ΔD, where π = operating surplus, W = employees compensation and D = depreciation. As D is determined by the quantity of  depreciable assets produced in prior periods, we have ΔD0. Consequently, when we make an economic, analytic equation with the economic model  shown in Fig 5-1, only the directly proportional equation between ΔY and ΔW emerges because π is directly proportional to Y in the chart.  [ This paragraph might be rewritten later. ]

However, W is in truth, expressed as follows: W = WV + WF, where WV = the component, in W, which is proportional to Y, and WF = the one, in W, which is not proportional to Y, and so W is a two variables function. In business fluctuations, WF is not a constant but a variable which is determined by a management executive's will according to business conditions. He changes both WV and WF in such a manner that π/Y is held constant or π0 is kept even in bad business conditions. In analyzing a business fluctuation problem, one must not assume ΔWF= 0. The analysis for the  unemployment problem is the very one for the ΔWF

In the break-even type chart shown in Fig. 5- 2, only the end point shows the realized GVA (input) and the realized output( products).  In the figure, GVA is expressed as income.

Fig 5-2 Break-even type chart

The characteristics of this chart are as follows:

(1) Any point on the intermediate position of the 45 degrees line connecting the origin to the end point shows an imaginary coordinate point( imaginary output, imaginary input).

(2) The operating surplus( 'Profit' in Fig.5-2) is a function of 3 variables, the whole final products Y2 , a variable cost ratio and fixed costs in GVA. If we take stocks into consideration in this type of chart, we can draw the 'break-even line' as described in §1 in 'Management Accounting'.

This break-even type chart includes the break-even point which is a result of the relationship between input and output. In business management, business entrepreneurs always manage the relationship between input( costs and profit) and output( sales) .

By the way, the shape of the graph in Fig 5- 2 always changes a little in an accounting period, because the expenses which are regarded as fixed costs at the beginning of the period, always fluctuates according to both business managers' and( for wages)  governmental( for depreciation) will.

As Y2 is placed at the same position in the both charts, the concepts of these two charts are prone to be misunderstood. However these two charts show different concepts.

The Keynesian's consumption function shown in Fig.3-3 is derived from the time-series type chart concept. What kind of information can we expect  for the additional government expenditure ΔG  from this chart? The answer will be that the yearly average proportion of ΔG( revised additional amount)  to G (amount on an initial budget) can be calculated. The income change corresponding to ΔG is, as a matter of course, the same amount as ΔG. However, the total income doesn't change mathematically, because ΔG is products which resulted from transfers between some people( or firms) and other people( or firms).

As the national accounts results from all firms' income statements, the national economy shows integrated individual firm's movement. This movement of the economy is Fig.3-13. The point K is the only point where profit=0. Both FV and FF are independent variables. This is the real economy. In the real economy, both the Walrasian equilibrium and the Keynesian multiplier effect can never exist.

Fig 3-13 Break-even type input-output chart

The incremental part enclosed with the point K and K' is the same as Fig.3-14, although the both appearances are a little different.

Fig. 3-14 Relationship between Components in Δ G

The description in this part (19) might concern being right or wrong( incorrect/ flawed) for the profit maximization equation, Δπ = ΔY2 - (ΔW( wages) + ΔD depreciation））=0, by the classical school( as used in Keynes’s theory ), which gives the first postulate wage equals marginal product of labor. The reason is as follows: (1) Costs including wages consist of both fixed and variable costs. The fixed costs are not constants but a kind of variable, which is determined by human will and changes without relating to final products quantity, whilst the variable costs change in proportion to the quantity; (2) The equation Δπ = 0 means π equals a constant, so that it may not be the profit maximization equation.

Within an accounting period when depreciation is almost regarded as constant, we have, from Eq.(3-40), Δπ = (1 - a V) ΔY - ΔWF, where a V = WV / Y ( variable wage cost ratio), ΔWF = incremental, fixed wage cost and ΔY = ΔWV + ΔWF. The equation Δπ / ΔY0 means that π is constant, so that the wage line is parallel to the 45 degrees line.  Consequently, the equation Δπ =0 is not the profit maximization equation but that the equation Y = W + π (const.)+ D( const.)  i.e. ΔY = ΔW is assumed. This is impossible depending on the break-even chart which shows a real management. In any age entrepreneurs pursue the realization of the largest profit margins or the minimum losses through improvement of human abilities and increase or decrease of the ΔWF.   It is of my current opinion that the profit maximization equation does not exist in production, even though ensuring of profit does. Therefore, the author is suspicious of the first postulate, so it is possible that my stance for this part may change after consideration. [The description of this paragraph is not final and will be corrected. Using the notation π(const.) might be incorrect. ]

Although both employees and employers are content with a wage level, the author agrees with Keynes’ deduction that an unemployment situation exists regardless of the content. That is to say, the labor forces in enterprises are not determined by a supply-demand relationship between employees and employers with a parameter of wage, but are actually determined at the crossing point between final products supply and effective demand.

See Fig 3-13 which expresses the break-even type Δ G-Y chart. In this figure, the above paragraph states that the labor forces which are included in the costs on the vertical axis, are determined in such a way that all the costs plus a profit are satisfied with the final product demand on the horizontal axis. Despite the erroneous assumption: Y = W + π( const.)+ D( const.) adopted in the theory, the correct deduction was reached over the assumption. Nowadays this is only to be expected. However, this deduction is amazing when we consider that the break-even type profit chart had not yet spread( and perhaps still hasn't even today) over the economics field in the days of the work "The General Theory of Employment, Interest and Money".

(20)   (1) When initial investment spending ΔI is implemented, investment demand rises by ΔI, which generates production by ΔI. The increase in production changes into income by ΔI according to GVA. (2) This income increase ΔI in turn raises consumption demand by MPC·ΔI,  which generates production by MPC·ΔI. This production changes into income by MPC·ΔI. (3) This income MPC·ΔI raises consumption, production and income by MPC2·ΔI as well. (4) This feedback continues indefinitely, which results in the multiplier effect : ΔY = ΔI + MPC·ΔI + MPC2·ΔI + · · · = ΔI / (1-MPC).

At each step, the following equations hold: increase in consumption demand = increase in consumption products  = increase in income from production. There appears to be no contradiction anywhere. This logic is that which has cemented the multiplier effect theory in modern economics. It is extremely difficult to simply explain the mathematical errors in this logic, so this has resulted in an extensive explanation.

Before J.M.Keynes, there had never been another person to consider this line of logic. Persons with no grounding in economics will not generally know of this logic. People in general tend to view economies as follows: products( as inventories) are manufactured depending on the quantity of labor, with labor being defined as employees, plants and equipment. They don't see a profit until a product is sold( really after a time period has passed).  In short, they think that no working or selling will result in no manufacture of products, hence no income. Many also believe that real economies are based on addition and subtraction. In fact, this belief is correct.

As is generally known, the logic of the multiplier effect uses the following Taylor series expansion:

 1/(1 - x) = 1 + x + x2 + · · ·             [ |x| < 1 ],    x = MPC (6-1)
 x/(1 - x) = x ( 1 + x + x2 + · · · ) (6-2)

The logic by Taylor series is reduced to the following problem, as shown in Fig. 6-1. Can initial investment expenditure ΔI (fixed value) generate automatically consumption ΔCCR  (variable value)? This question is by nature the same as,  can we  automatically get ΔYCR = ΔCCR/ MPC, if we give initial consumption ΔCCR ?

Fig.  6-1

First, we mention about whether the multiplier effect equation ΔY1 = ΔG2 / ( 1 - MPC)  by  government investment expenditure ΔG2 holds. As is explained in (10), since the Keynesian cross condition doesn't hold in this equation, this is a mathematically wrong (incorrect/ flawed) equation. Consequently, a measure against unemployment by a government due to this equation is mathematically meaningless , so from now on, this equation will not be considered

The problem is the equation ΔY1 =  ΔIT / (1 - MPC) which appears in Keynes' original chart shown in Fig 5-5. ΔIT is highly correlated with private investment.

Fig 5-5 Keynes’ original chart

This equation claims ΔY2 = ΔIT / (1 - MPC) because ΔY1 = ΔY2. This equation is mathematically correct  as a relational expression between the symbols. However, this relationship only holds  when consumption products ΔC2 are supplied and purchased, continuously and simultaneously, keeping the proportion MPC ( = ΔC2/ ΔY2). It is understood, from Fig 5-5, that we eventually acquire only income, ΔIT,   which is equivalent to the initial investment expenditure ΔIT.

So this raises the question, how should the sentences in the first paragraph be rewritten? The first sentence "(1) When initial investment spending ΔI · · · · according to GVA." means that ΔI2 (product by firms and its purchase by others) = ΔI3 (cost and profit in firms) = ΔI1 ( ΔGVA corresponding to ΔI3 ). This is correct as it stands now.

For the sake of decreasing symbols, we will not consider government finance here. This operation doesn't give logical inconsistency because the ΔG2 problem is not considered here(ΔG2 = 0 ), and the multiplier effect problem is the one between incremental components of GVA. What are the resources( inputs) of ΔI ? Those are capital depreciation ΔD (=0) and saving ΔS = ΔπI + ΔπC + ΔWI + ΔWC - ΔC, where π= profit, W = wages; subscripts I and C show that the quantity with each symbol belongs to the symbol products ( f.e. C).

In fact, when we say  "initial investment spending ΔI is implemented", the sentence "only initial investment product ΔI is produced, while consumption product ΔC is not produced" is requested as an implicit premise. Then,  we have ΔC = ΔπC = ΔWC = 0. This results in ΔS = ΔπI + ΔWI.  Under this premise, as a matter of course,  consumption doesn't increase because all the incremental incomes( = incremental wages + incremental profits) are expended in order to produce ΔI. Consequently, if we rewrite the first sentence as " only the initial investment products ΔI is produced and sold", the meaning of the latter sentence  becomes clear, and so the sentences (2), (3) and (4), which follow the first one, don't hold.

Although the author has stated, at the preceding description, that the sentence "only initial investment products ΔI is · · · as an implicit premise " is requested, Keynes' economic system, in the nature of the theory, can't allow the request. In Keynes' system, ΔC2 with ΔIT is one pair, and they can not be separated.

So in this case, the first paragraph should be rewritten as follows: if we ensure simultaneously the implementation of additional private investment( production and purchase) ΔIT and additional private consumption ( production and sales) ΔC2, at the proportion of MPC to 1 - MPC, in a certain time period, we can get additional inputs ΔC1 = ΔC2 and ΔI1 = ΔIT at the same proportion. The sentences regarding the multiplier effect (2), (3) and (4) don't follow the first sentence (1). If we could execute the measures of the first sentence , Keynes' economic system would be possible. However, even so, the multiplier effect doesn't exist.

That is to say, the above-mentioned clause 'When initial investment spending ΔI is implemented,' is not compatible with Keynes' multiplier effect theory. If an economic measure is like this, the MPC condition is not kept constant. If we change the clause into ' When additional investment and consumption ΔI + ΔC are  implemented,' the multiplier effect doesn't exist.

Thus, we arrive at the conclusion that all the sentences in the first paragraph, '(1) When initial investment spending ΔI is implemented,  · · · (4)  · · ·  = ΔI / (1- MPC).' do not hold as far as the MPC condition is preserved.

Let us return to the first. Is the following proposition correct?: if a government pay out cash benefits (fixed amount handout) ΔCMONEY to people in order to stimulate consumption, investment fund ΔIMONEY corresponding to a marginal propensity to save ( tax=0) will be induced. As a result, capital goods ΔI(=ΔIMONEY) will be produced apart from the production of consumption goods ΔC=ΔCMONEY. This does not hold in a depression. The reason is that the kinds of the consumption goods ΔC and the capital goods ΔI are different from each other. In a depression, plant and equipment including workers for both consumption goods and capital goods are in overcapacity conditions. Therefore, even if we assume that the cash benefits do not change into savings in the relevant year, although ΔCMONEY can be transformed into the consumption goods ΔC, capital goods ΔI can not be produced. In addition, if consumption falls as much as ΔCMONEY in the next year, the borrowed money ΔCMONEY will remain.

Let us return to the first. Is the following proposition correct?: if a government pay out cash benefits (fixed amount handout) ΔCMONEY to people in order to stimulate consumption, investment fund ΔIMONEY corresponding to a marginal propensity to save ( tax=0) will be induced. As a result, capital goods ΔI(=ΔIMONEY) will be produced apart from the production of consumption goods ΔC=ΔCMONEY. This does not hold in a depression. The reason is that the kinds of the consumption goods ΔC and the capital goods ΔI are different from each other. In a depression, plant and equipment including workers for both consumption goods and capital goods are in overcapacity conditions. Therefore, even if we assume that the cash benefits do not change into savings in the relevant year, although ΔCMONEY can be transformed into the consumption goods ΔC, capital goods ΔI can not be produced. In addition, if consumption falls as much as ΔCMONEY in the next year, the borrowed money ΔCMONEY will remain.

The logic of this part is very important. Consider an economic condition where: (a) it is in near full employment;  (b) plant and equipment are not in overcapacity; (c) business conditions are averagely on a profitable basis; (d) public capital goods are effectively useful. (e) assets held by financial institutions are sound. In the situation, an autonomous increase ΔC of private consumption goods can be inducement of a new investment fund ΔIMONEY. If banks supply creative funds ΔIMONEY to enterprises, ΔIMONEY will be transformed into capital goods ΔI. Further, if wages are increased as much amount as maintaining the marginal propensity to consume being constant simultaneously around or immediately after the consumption, GDP will increase. If the wages are not increased, the facilities will be immediately in overcapacity. When many job seekers of low wages exist, another phenomenon will be produced.

The economic processes of very good and very bad business conditions do not become a reversible process. The reason is that the preconditions described above are quite different from each other in both business conditions, so that  each different economic phenomenon occurs. That is to say, each economic phenomenon is irreversible one.

(21)  Commencing with Paul Samuelson’s multiplier–accelerator macro dynamic model*, several papers which developed the simple Keynes multiplier effect model, were presented. Through those papers, the above-mentioned treatment (20) holds. To simplify matters, the following equation is used.

 ｚ（ｔ）＝ｘ（ｔ）+ｙ（ｔ） （6-3）

where time is denoted by t, Cconsumptionby , Iinvestmentby y, and GDP by z.

In this equation we have three variables , y, and z, where any of two can be independent variables and the remainder is a dependent variable.

 ｘ（ｔ）＝a · ｚ（ｔ - τ） (6-4)

, where a is a constant and τ is a parameter which expresses a time lag.

If we add Eq.(6-4) to Eq.(6-3), the two independent variables automatically decrease to one. This results in one fixed or unique pattern of relationship between (t), y(t) and z(t), although the variables are indeterminate with respect to time and can move freely in relation to time.

If we add the following equation:

 ｙ（ｔ） = ｂ·（ｘ（ｔ） - ｘ（ｔ - τ）） (6-5)

, the movement of all variables becomes fixed and has a certain pattern with respect to time. Any form of equation such as a differential equation is allowable in place of Eq.(6-3) and Eq.(6-4), when the same treatment holds.

The most important point is that Eq.(6-3) is always satisfied by the solutions of any two equations at any time by the definition of the economic terms. Eq.(6-3) shows that(t) and  y(t) are components of z(t), and the three volumes of them are already generated in a unique form at the same time. That is to say, if we don’t have an economic means that realizes (t), y(t) and z(t) surely and simultaneously, the economic model will not be able to be used as an economic policy. Solutions obtained by any three equations with initial conditions will be fixed to time, so economic movements after the initial conditions will have be determined. Given this solutions, any economic policy after the initial conditions will no longer be effective. In short, since the investment multiplier effect itself does not exist, these analytical methods are nonsignificant.

* Paul A. Samuelson, THE HISTORY OF ECONOMIC THOUGHT, WEBSITE, http://cepa.newschool.edu/het/home.htm

(22)   We shall consider the case where private-sector demand decreases. If we can increase government investment based on its economic policy for the decrease, the government investment will certainly be replaced with it even though the multiplier effect doesn’t exist.

We can try to imagine a nationalization process of private enterprises to state-owned ones. We assume, as a mater of course, that the concepts such as depreciation, sales and profit in the latter exist. The unemployment due to the decrease of the private-sector demand will not occur, if the following situations arise: Nonexistence of government borrowing increase; no competition against other countries and private-sector enterprises; ensuring sales by forcing; no business responsibility for losses.

If this process is executed with an increase of government borrowing, the greatest obstacle to the unemployment relief is the absence of the business responsibility for losses. This is because we can’t clearly separate profit (loss) from depreciation recovered by sales. This can be understood from the following: On the one hand, in a piece of  equipment, a decrease of value due to physical deterioration as well as obsolescence i.e. a reduction in operating capacity is caused. On the other hand, a recovery of investment fund (cash purchase amount) is promoted. These two money values don’t necessarily coincide with each other in each accounting period. The loss corresponds to the impossibility of recovering the initial fund.

If, through the government economic policy, the state-owned enterprises run losses for years, then the previous increase of government borrowing can be returned only by a new refinance. The point of this logic (the problem of the business responsibility for losses) is that the person, who can receive the reward from the capacity i.e. the recovery of the equipment expense by means of product sales, should be the owner of the equipment. In short, his results from the following: nongovernmental enterprises consider that the depreciation in business accounting through sales is a means of recovering of an investment cost, while government officials underneath think that the consumption of fixed capital( if the term depreciation is used) in social accounting is the cost for an imaginary government final consumption  (remaining value of capital goods with ages is pointless in accounting.).

(23)The economy in the one-degree of freedom system means that if people in an economic field moves toward a direction, other parties in the other field will automatically follows to the same direction. In Keynes’s multiplier effect theory, the generation of the final products = consumption goods + investment goodsshould be in the one-degree of freedom system, because the MPC really exists statistically. Consequently, both consumption and investment should be generated simultaneously.

However, such a movement between consumption and investment in the one- degree of freedom system naturally occurs in a stable economy with a few changes in economic growth, aside from the multiplier effect. The reason is that final products are equal to incomes. Relevant parties of consumption and investment (even in this field buyers and sellers exist) observe each other at any time. If one side convinces the cause of the other side’s activity, the former follows the latter.

The following phenomenon possibly occurs: with a stable economic condition in a short run, an export ΔEX suddenly increases rapidly; a national income increases by just that much; a new domestic consumption is increasingly produced after the stimulation by the consumption; the domestic investment as 1-MPC/MPC times of the consumption is executed; a consumption responds to the wages included in the investment again. In this manner, although the investment multiplier effect doesn't exist, we can say that a stimulation effect of investment to consumption, or the reverse might exist, but it doesn't always exist.

Therefore, the author doesn’t deny the economic policy based on the principle of effective demand by Keynes because of my conclusion that the investment multiplier effect doesn’t exist. The government must actively address the national measures that cannot be achieved by individuals and the private sector in the system without an increase of public bonds. If the government doesn’t wish to undertake these economic measures, then that government has no place to exist.

(24)   Even if anyone looks at Keynse's investment multiplier effect formula or examines it, the formula appears to be surely an identical equation. Even the author sometimes lose the reason of the formula's error after a little time's passing. Important parts to know the error are: the item (10) in this section; in Fig. 9.3, §9, the correct expression, 5+ΔG= Y** = ( 2+ΔG)/1-3/5+ΔG））, giving ΔG=ΔG, presented by the author, on the other hand, the wrong expression, ΔG ΔG/(1 - 3/5)=2.5ΔG, giving 1≠2.5, derived from the equilibrium theory. Additionally refer to the writing under Eq.(19) in Section 4, in the paper, Yuichiro, Hayashi, “Production Theory to Analyze Human Wisdom and Global Phenomena Operations”, on this website. Only a little mention concerning this problem is made.

Conclusion

The error in the Keynesian multiplier effect theory results from the following 4 fundamental mistakes:

(a)   One is a misinterpretation of the flowing of partial GVA, including transfers( see 3. taxes problem in §5) in the three aspects, Y1, Y2,and Y3, within a year. If we adopt the assumption, Δ I2 = ΔNX2 = 0, the partial output amount of Δ I2 + ΔNX2 is taken away from the total ΔY2 , although ΔY1 holds the total input. That is to say, the Keynesian cross condition ΔY1= ΔY2 is not satisfied.

(b)   The second is confused understanding of the two problems with the two different  assumptions; (i) ΔI2 = ΔNX2 = 0 and (ii) ΔG2 = ΔNX2 = 0.

(c)   The third is the fact that ΔC2, which has been realized simultaneously with the execution of ΔG2 ( or correctly ΔI2 ), and is hidden from sight in  the Keynesian multiplier effect chart( Fig.3-3), has been overlooked. On this point, the author will consider the cause included in the equilibrium theory.

(d)   The fourth is as follows: Keynes' original chart model is a two-degrees of freedom system in which variables are both C2 and I2, and Y2 is a dependent variable. Keynes didn't know that the system created a one-degree of freedom system just when he assumed the existence of the MPC. In this system, the amount ΔIT ∙ MPC / (1 - MPC) = ΔC2 must be provided simultaneously with the production ΔIT in order to keep the MPC condition. When both ΔC2 and ΔI2 are fixedly connected to be an economic quantity with a one-degree of freedom, the series multiplier logic of production (= income) dose not hold.

Incidentally, Fig.4-7 is correct as a time series type input-output chart. However, Fig.4-7 has been made only for explanation, so such an economic phenomenon( independent increasing of only ΔG2) will not occur in real-life. Fig.3-13 is correct as a break-even type input-output chart.

Fig.4-7 Correct time series type input-output chart

When we look at final account data in  both a time-series type chart and a break-even type chart at the end point of one year, both data sets are the same. However, the former chart shows only data as a final result, but the latter chart shows the cause of the final result, as well as measures to manage the economy, in addition to the final data. When we want to analyze economic problems, including those of  unemployment, we must use the concept which is implied in the break-even type input-output chart.

The errors in Keynes' multiplier effect theory resulted from several errors overlapping each other. Each error was not based on small mistakes but that relating to the fundamentals of conventional methods for analyzing economic problems. If readers don't advance step by step by checking the above-mentioned observations (1) - (21), understanding the error of the Keynesian theory as a whole may be difficult.

The end of the disproof

Postscript

A part of descriptions has been deleted.

In fact, there are two main causes for the result of the huge cumulative government  bonds in Japan. The one is as follows: Keynes's principle of effective demand theory, based upon the investment multiplier effect, is originally the theory which gives relief to the unemployment by a government; whereas, in Japan, this theory has been used as a basic theory grounding the means since a year for (1) economic  development, (2) overcoming a depression and (3) expanding domestic demand, going beyond its purpose; surprisingly, the basic formula  grounding this principle is itself wrong( incorrect/ flawed)! However, the error of the formula is the same all over the world. This abnormal state in Japan doesn't result from a mistaken economic policy such as applying a theory to a field beyond its application range, but comes from the Japanese national character of a single-mindedness that if people accept a thought, they never allow discussions about different views from various sectors nor thinking including disbelief.

On this subject, the followings are concerned:  Japanese remedies for the depressions resulted from the adoption of the floating exchange system by the U.S. in 1971 and the first oil shocks of 1973; the serious trade friction between Japan and the U.S., especially due to the rapid increase of automobile export from Japan to the .U.S. in 1980s; the announcement1990 of public investments worth 430 billion yen over a ten-year period starting in 1991 for domestic demand expansion ; antidepression policy, with public investments, for handling the big depression occurred by the burst of the bubble economy at the start of 1991.

By the way, the government public investments expenditure of 2006 already went back to that of before the birth of the bubble economy. Consequently, the economic crisis problem due to the erroneous Keynes's investment multiplier effect formula is already over, apart from the repayment problem of the principal national funds.

The other is as follows: the yearly social security costs in Japan, which are a typical consumption expenditure ( value decomposition of goods and services produced in a given year), has not been kept within the limit of the present tax revenue; then the deficit became consuming, of cumulative national savings, which can't be expected to recover (recovering by means of sales) in the future.

This mistake doesn't originally relate to the advent of the aging society in Japan, but this results from a mistaken economics thought and political measures based on it, which have continued from the past. This is unrelated to Keynes’ multiplier effect theory but this results from a big mistake in economics.  Politicians mistakes are secondary, because politicians' economic understanding, including that of bureaucrats and media, doesn't surpass that of economist.

Why has the mistaken thought, that people in the present and future generations should bear the senior welfare and pension costs of the past generation, endured and been allowed in economics for so many years? It is abdication of learning that if one might attribute its reason to the historical origin of the establishment of the pension system. Why haven’t economists advocated the logic that the value of human life must not be so high as to threaten the people’s livelihood ( i.e. life) both in the present and the future?

According to the Keynesian’s systems of 3-accounts in the national economic accounting, the present yearly investment by people, including corporations, is only due to the present yearly saving (gross) by the people. Only the people that invested in the past receive the present yearly return of investments (repayment of debts). All the people, who didn’t invest in the past, accept no return in the present. Medical spending for human life has the same economic meaning as military spending for the fate of the nation. In addition, no creature has such a principle to ensure their survival into the future, and gives priority to their life rather than that of their offspring. Economists have responsibility for pursuit and argument of truth in economics. They need not accommodate themselves to the wish of the public.

Concerning the foregoing descriptions, refer to the following website:　Current Japanese Fiscal Conditions and Issues to be Considered (2003)

Let’s get back to the point. Anyone could give such easy explanations or questions from the beginning of Keynes’ work. Keynes himself understood the explanation more than anybody. At that time, all but he gave such explanations. In this economics situation, Keynes struggled to resolve the unemployment problem to find the revolutionary logic: a certain volume of investment generates the income of the investment multiplier times that of the investment; involuntary unemployment is generated through enterprises’ adjustment of the labor quantity due to inflexibility in labor prices; in order to rescue unemployment, public investment must be used to become the effective demand. Thus, all the economic minds around the world have admitted that the logic is ‘mathematically true. From that point, nobody has been able to disprove the logic.

Joining the word ‘mathematical’ to the sentence ‘Anyone can reproduce a phenomenon based on a hypothesis ’ means ‘scientific’. Keynes’ multiplier effect theory has been ‘almost scientific’ for many years. Therefore, this theory is, even now, placed in almost every textbook of economics.  Thus, this theory is taught at a primary stage not because it doesn’t fit the economic reality or it is an old theory but because it is the most important foundation, supporting the modern economics, which has been conceded as mathematically true. The textbooks in which the multiplier effect theory is not placed, take precedence of the fact that the theory doesn’t fit the economic reality but they skirt the mathematical disproof of the theory. How do they explain the relationship between income and demand?

At this moment, any economist knows that the investment multiplier effect doesn’t fit the economic reality. However he does not dare to express a completely negative view against the multiplier effect theory based on a mathematical proof at the risk of his career as an economist.

In order to release the economy from the influence of the multiplier effect, it is imperative to disprove mathematically the multiplier effect theory, and to present a new theoretical method to analyze the unemployment problem in place of the multiplier effect theory. The reason is that firstly only the disproof of Keynes’ multiplier effect theory without presenting a new analytical method in place of it will return nowadays to the situation 70 years ago.

Secondly, such things as the economic fluctuations, instinctive economic remarks and the entire ordeal which followed the economic bubble burst in Japan, will have been forgotten a few generations later. Each economic measure executed in these years without the support of mathematical logic will be treated as one of many feasible measures in order to recover the result due to a failed economic policy. Still, after the years, remaining logic will be that that can still be admitted scientifically or mathematically true even in that year. The author can’t imagine how damages might be produced to the world economy, if the economic situation in the 1990s in Japan would reoccur in China or India.

Keynes’ multiplier effect theory is not one that can be completely denied with simple explanations. No, the possible existence of the investment multiplier will not be able to be denied entirely in a certain economic situation. The reason is that the multiplier is like a fine thread (as the economics situation now stands), which connects production and demand. Thus, the MPC or the investment multiplier would remain in economic models into the future as a coefficient that is largely associated with economic group propensity or business executives’ will.

A few courageous economists *2 have denied for the multiplier effect theory at the risk of their careers as economists despite the cold reception and criticisms of the academic world, even though they couldn’t give mathematical disproof. The aforementioned behavior is undesirable with the view that it was not accompanied with a disproof on par with Keynes’ proof for the multiplier effect; alternatively, cited references of past nay-sayers was not given.

Moreover, this behavior is discourteous to the following: Keynes himself; other famous economists; economists who have devoted their whole lives to the Keynesian theories; the preceding nay-sayers; university students over the world who are taught the theory even now; unemployed people, bankrupts, suicide victims and business managers in unfavorable business conditions. Incidentally, the recent sales of the authors’ company fell by a 1/3 in comparison to the previous average sales for several years.

It has taken 11 years to reach here from the sudden idea of the managed gross profit chart theory. The author’s paper on that theory was presented to proceedings of a Japanese accounting academy several years ago. The original copy was ordered to be  largely rewritten more than once or twice. Its publish in the proceedings was finally rejected forever because its reading referees and the editorial board couldn’t ascertain whether the author’s solution, or Solomons’ solution appearing in accounting textbooks,   was true for the same accounting problem. Through constructing the disproof logic for Keynes’ multiplier effect theory, the author experienced, a number of times, feelings of hopelessness, because immediately after he thought his question had been solved, extremely difficult problems emerged one after another. The author himself has made all this effort.

Keynes is still the greatest economist in history through the creation of the principle of effective demand based on the investment multiplier effect theory and other outstanding achievements. Who but he could write the work ‘The General Theory of Employment, Interest and Money‘ and construct the fundamental framework of modern economics?

Readers might think strange that the author has cited no earlier references for the multiplier effect theory throughout his formulation of the disproof. We have, in fact, both many research papers that intended to develop Keynes’ original theory, and a substantial number that intended to deny intuitively the multiplier effect. The latter references admit the existence of the investment multiplier, and state that the effect has declined for some reason or other. However, as far as the author knows, no study has existed, whose primary purpose is to mathematically disprove the logic. If such a study were to exist, and completely succeeded in its purpose, the multiplier effect theory would not appear in economics textbooks. Furthermore, such a large postwar economic growth in Japan and the huge quantities of national bonds (plus local bonds) might not have been realized.

I went back and forth among many equations and charts correcting them. This web site shows my thought processes. I now find no small number of inconsistencies or duplicate descriptions within the sections. However, if I were to revise my descriptions consistently throughout, I would need much time. Such work will not change my conclusion that the Keynesian multiplier effect simply doesn't exist.

In closing, I don't say that we don't need public investment but say that the Keynesian multiplier effect of public investment doesn't exist both mathematically and in accounting formulation. A public investment might trigger an upturn or recovery in the economy. However, the investment  might cause such an effect  but it will never produce the multiplier effect.

The proper public investments are indispensable both for social rearrangement of incomes( or products) and economic growth.

Economically sound investments are the solid foundations for the people's future.

Excessive consumption due to public funds is a malady that will eventually result in the economic demise of the nation.

Slight modification  2013, Aug. , 25

Research process