The Kid Who Destroyed Keynesianism

That’s right, a 21 years old boy destroyed keynesianism long ago. In 1947 Murray Newton Rothbard, born in 1926, wrote an amazingly good and direct critique of keynesianism. Some of the arguments he used were not his own creation, but others certainly were. Despite this the article is brilliant. Here is young Rothbard vs The Big "K". Let's see what he thought.

"K"eynesianism working

These are the two fundamental pilars of the elemental keynesian model.

1) The constant consumption and investment functions 

Let's start with the basic keynesian 101:

Aggregate Income = Aggregate Expenditures

You know, this means that “your spending is my income” and all that...

Let’s put an example:

E = Y = C + I

C = 0.9Y

Y = 0.9Y + I

We see that the 0.90 is fixed, so consumption (C) is passively determined just by income (Y). The “pasive” factor is 0.90, usually called “marginal propensity to consume”, thus consumption is determined by income. 

However investment (I) is not determined by Y. Whatever determined it, was not important (Animal spirits? Expectations? Interest rates?). The important thing is that it was not income (Y).

All we have to do now is to use basic math to isolate Y:

Y = 0.9Y + I -> Y - 0.9Y = I -> 0.1Y = I -> Y = 10I

Income is equal to 10 times investment or in other words income is 10 times the element which is not dependent of  Y.  

If we include goverment in this model, then we must include government expenditure or G.

Y = C + I + G

Y = 0.9Y + I + G

Now, government G has two features: 1) +G means a deficit and -G is a superavit and 2) G is also independent from Y. Notice that, from the point of view of this model, private investment (I) and government spending (G) are both independent or "autonomous" from Y, and both have the same economic effects. Assume Y = 100, then C = 90, S = 10 and I = 10. Where S = savings and there is a balanced-budget government (G = 0, there is neither deficit nor superavit).

100 = 90 + 10

In this "equilibrium" the aggregate of firms and consumers are satisfied. Now government has a deficit of 10 or investment increases in 10, it is exactly the same, so we get: 110 = 90 + 20.

Now I = 20, why Y is not 200? Because the above equation is a disequilibrium situation, the automatic-multiplier process is not over yet. Now with that income 110, consumers will spend 99 in consumption and will save 11. Notice that investment is 20 and saving is 11 right now, so I>S and the economy (Y) is expanding (from 100 to 110). But is not over yet: Business are happy seeing people spending 99 instead of 90, so firms pay income 99 + 20 = 119. And now again consumers receive 119 income so they consume 107 and save 12. Firms are even more happy because they expected 99 and they are getting 107, so they pay 107 + 20 = 127. The process continues and ends when Y = 200, C = 180, I = 20 and S = 20. Finally S=I, income has grew enough to make saving equal to investment.

200 = 180 + 20

Rothbard says: "It is important to notice that equilibrium was reached in both cases when aggregate investment = aggregate saving... When investment is greater than saving, the economy expands and national income rises until aggregate saving equals aggregate investment. Similarly, the economy contracts if investment is less than saving, until they are again equal"

Here is the key for this process to work: "two very important things must remain constant in order that equilibrium be reached.":

a) The consumption function (and therefore the savings function) 

b) The level of investment (independent from income)

2) The level of income as determining the volume of employment

Assuming all this constant: a) technology, b) efficiency, quantity and distribution of all labor, c) quantity and quality of capital, d) distribution of income, e) relative prices, f) wage rates, g) consumer tastes, h) natural resources, i) economic and political institutions... Then!, we can say that to every level of national money income there corresponds some definite and unique level of employment. The higher the money income the higher the level of employment. But income can be at equilibrium at a level not corresponding to "full employment" and according to keynesians, there is no reason to assume that the level of income determined in market economy coincides with the volume of employment considered as the "full-employment" level.

"K"riticism

1) The constant consumption and investment functions

a) The static problem: The aggregate method hides the fact that they cannot remain constant. The keynesian model fails because uses aggregates that are meaningful in the arithmetic world to operate, not in the real world. The aggregative construction of the model made us believe that in some equilibrium situation the aggregate of firms and consumers are satisfied because their expectations are fulfilled. The fact that a single firm receives just what expected does not mean that it is in equilibrium. Inside the aggregate there are firms losing money and other making great profits, others making moderate profits and great loses. Even assuming that the loses of one group cancel the profits of other group, every single firm will make its own adjustments investing more or less according to their valuations. The aggregative form of the model hides all this. Investment cannot remain at 10, it cannot remain constant and the consumption function also will not be constant, all this will change the level of income. The aggregate nature of model does not tell us in what direction or how far this variables will move.

b) The dynamic problem: They cannot remain constant when economy is expanding. Even assuming the first criticism is wrong and both can be constant in the static state, they cannot remain constant in the required time in which economy is expanding. When aggregate investment is greater than aggregate savings, economy expands to a level of income where they are equal. But in the very process of expansion, consumption function (and savings) and investment cannot remain constant. In an expanding economy there will be profits that will be distributed among firms in an unknow way making necessary some adjustments and changing in investment, the model of course says nothing about this. And new firms will enter the economy in expansion, changing again the level of investment. But besides, the distribution of income in an expanding economy also necessary changes and so it does the consumption function, which will change in an unknown direction and magnitud. The model does not say anything about that changes. Capital gains in firms of the expanding economy also changes consumption function.

c) Indeterminacy: As both are not constant, the level of income in the system is indetermined. Since the two basic determinants of income, the consumption function and investment, cannot be constant, they cannot determine any level of income, not even approximately. "There is no point toward which income will move or at which it will tend to remain. All we can say is that there will be a complex movement in the variables of an unknown direction and degree." Not only the model fails, it is also useless. 

d) Aggregation fail: The model fails due to its aggregation level denying individual things. The aggregative concepts of keynesian model are misleading because leaves out other important individual things. Consumption is not just a function of income, it also depends on past income, expected income, the phase of business cycle, lenght of time, prices of goods, capital gains or losses, cash balances, etc.

e) Ignores interdependency of the economic system: By splitting into a few aggregates, the model assumes they are independent from each other, that are determined independently and that they change independently. Saving is not independent from investment, firms save in anticipation of future investment. Anticipations of profitable investments influence savings function and hence consumption function. Investment is also influenced by level of income, the expected level of future income, anticpation of consumption, by the flow of savings, etc. The assumption that state just adds or substracts income by deficits or superavists is also false because it ignores interdependence between aggregates. Investment will not remain constant, if state tax to encourage consumption then savings will fall affecting investment. If state tax profits, that will also modifies investment. Consumption and investment are determined by complex interactions between and within them.

f) Statistical problem: There is no stable consumption function. Keynesians have been, as far as 1947, completely unsuccessful in proving statistically a stable consumption function. Consumption changes with the month of the year, the business cycle, in the long run, habits, family income, expected changes in income, etc. This is the final nail in the coffin for the income determination of keynesian system.

2) The relationship between income and employment

a) Hyper-short-run assumption: The constancy assumptions needed for this relation to exist are only possible in a very short period of time. The only way for all that variables to be constant or "given" is during a fraction of a second. During that minimum amount of time, they are certainly "given". During a millisecond it is absolutly true that all that variables are constant. 

b) Extending short-run assumptions to the long-run: Keynesians use this very short-run relation to predict. They use this model based on very short-run and unreal assumptions on periods of time far longer than them and that makes them fail. Like keynesians predicting a very hight unemployment after WWII. 

c) Money wage rates constancy: The essence of keynesian theory rests on this assumption. The keynesian "underemployment equilibrium" depends on this. Now the fact that if a price does not fall enough to "clear" the market was neither new nor original. And if wages do not fall due to state privilege unions then it is the state what is creating "underemployment equilibrium", not the market. So the basic keynesian tenet that nothing in free market assures full employment equilibrium collapses.

d) Confusing wage rates and wage income: Keynesians think that wage cuts reduce spending because of this confusion. When keynesians think about wage cuts, they do not see that when this happens hours worked and quantity hired also increase (varying from industry to industry). So total payroll increases, and also demand. Keynesians confuse wage rates (wage per hour) with wage income (wage rate multiplied by time in which wage is earned).

"K"onclusion

We saw that young Rothbard already knew what later would became one of the most demolishing criticism of keynesianism ever made. In MES he updated and improved above criticisms, but it is amazing that such a young boy could have destroyed what was the paradigm of his own Age. 

NOTE: In a very interesting footnote, Rothbard shows that at his 20's, he was not an anarchist yet: “This does not imply that democracy is evil. It means that democracy should be considered as a desirable technique for choosing rulers competitively, so long as the power of these rulers is strictly limited.”