An Engineer's Critique of Keynesianism

Though Keynes raised many good questions, and made a critical observation regarding wage formation which I've adopted, I will argue he ultimately fumbled at the one yard line. Complicating such a claim is the fact that there are numerous interpretations as to what Keynes "really really" meant. As a result, I will restrict myself to a couple core concepts that are generally treated as traditional Keynesian.

In a nutshell, Keynes argued that the consumer does not spend everything he earns. A more sophisticated form of this point is reflected in a concept of marginal propensity to consume (MPC), the details of which we will not address here. Unless offset by some other form of spending (e.g. business investment), Keynes argued it would leave a gap in the total aggregate demand in an economy resulting in unemployment.

In what was called Classical economics, it was believed long run unemployment was impossible, because Savings always equaled Investment (S=I). It was the rise and fall of interest rates which assured this condition. Keynes rejected this claim and thus concluded that there was no force driving in an economy to full employment. In contrast to what you might expect for a theory of recession economics, Keynes viewed consumer spending as stable, while business investment as the unpredictable component in aggregate demand due to fickle business expectations and interest rate dynamics. Thus to bring an economy to full employment (make S=I), investment would have to be spurred on one way or another.

This is Keynesianism in its simple form. It is sufficient to expose its fundamental errors. The first mistake lies with consumer behavior. Note the propensity to save was not something that occurred after a real estate bust or some other wealth destruction, but was simply normal behavior of the population as a whole even in the best of times. Unfortunately for Keynes, it turns out this is not the case during normal economic times. Instead, some consumers spend out of savings, while other add to there savings. As a nation, the net effect is zero on the growth of savings. In other words, on average the nation does spend everything it earns. This critique is not unique to me, and generally acknowledged as correct. One simply needs to look at the lack of annual national savings to see this.

Where it gets more interesting for me as an engineer is on the investment side of the argument. Notice that Keynes does not actually reject the S=I argument as a condition for full employment. This is critical. As an engineer who has argued there has never been a sound theory of economics (see link below), the S=I claim adds an interesting twist to the story because it not uncommon to find authorities who argue a nation needs to save to grow. This is nonsense, because it is a growth in savings that cause recessions, not growth. By this flawed logic if we saved every penny we earned we'd be booming. The reality on the other hand are banks which create credit effectively out of thin air for industrial investment. No savings required. This in turn begs the question what determines the level of investment under normal economic conditions. The answer, which Keynes failed to grasp in my view, is the state of technology. In other words, it is not hard to imagine that in the 1800s investment might make up 25% of GDP, while in 1900s it might only be 15% due to the advanced state of technology. To help appreciate this possibility, imagine a factory robot that might cost $100,000 today, may only cost $10,000 ten years from now.

From a bird's eye view, what this means is that if during healthy economic times investment say makes up 15% of the economy it will also be 15% during a recession. For example, if 100 Million people are employed, verse 300 Million it won't change the level of technology needed to maintain either level. This is of course an over simplification, but it emphasizes the point that investment is not a function of the marginal efficiency of capital and associated interest rates as Keynes suggested. In reality, you will make the investment level needed to produce the lowest cost product.

What this means in the end is that an economy can be fully functional with zero savings, and an investment level defined by the state of technology. S=I is dead. Once this is understood, it leads to additional complications for using Keynes as means to fix an economy. What Keynes also failed to consider was the the dispersion of investment across sectors. To see this, begin by imagining 100 million different buying patterns resulting from 100 million consumers. One consumer may consider books a luxury item, while another a ticket to a movie. The challenge this presents during a recession is that each consumer pulls back on purchases in his own luxury sector. Thus unemployment grows in small amounts in each sector, say 10 people at every business. This means, no government-driven road project can restore the resulting decline in 100,000s of businesses. Yet, it could be argued that a tax cut to each consumer might bring him to reconsider returning to his normal spending patterns. But the effectiveness of such approach will be a function of the wealth loss of the individual (following a bust) verses the size of the tax cut. In other words, if you just lost $100,000 on the value of your home, will a $1000 tax cut get you start spending again? Probably not for several years.

Adding to the Keynesian collection of errors is another defective concept known as the multiplier. The idea here is that one new dollar in new investment will "multiply" into many more dollars in the economy. It's theoretical root lies in the marginal propensity to consume mentioned above. The math is so defective that if no one was to save, the multiplier goes to infinity. Clearly something is wrong with a model that allows $1 to become an infinite amount when no one saves.

My interpretation rejects the multiplier on simpler grounds. Instead, I suggest that the multiplier cannot exist in any sense of the word, because an individual can only exchange eight hours of labor for someone else' eight hours of labor. My eight hours cannot multiply into fifty hours of labor. Consider a butcher and baker who exchange the output of their 40 hours of weekly labor. Introduce an unemployed road builder. Now let the government put the road builder to work. To pay him, we tax the bucther and baker. The butcher and baker eat less but a get a road for it, and the road worker now buys some of bread and meat. There is no multiplier here, just a reallocation of labor hour output.

What ultimately is the consequence of such a critique? It means we find our economy destroyed by free trade and under the false hope that Keynesianism can restore us. Though Keynes understood free trade was not a win-win game (he proposed the so called Bancor as a balanced trade scheme) and that imports were job destroyers, his focus was in stabilizing what he saw as an inherently unstable capitalism (like Post Keynesians today). But if both his assumptions about S and I are wrong, then his assumptions about how to fix an economy will not work. Ultimately sky high tariffs will be required. Lincolnomics, not Keynes' economics will prove the proper course of action.

As always, critique welcome.

The logic behind a new theory of economics:

Van Geldstone



This is ridiculous

This is just your philosophy and has little to do with real economics. A long rambling with a few concepts thrown in to try to claim Keynes isn't valid which most of it clearly is. Enough!

You must have Javascript enabled to use this form.

Not quite so

The point is that Keynes said that the adjustment of I and S is not done by the rate of interest (that was the marginalist or neoclassical school's loanable funds theory of interest of Wicksell and Marshall, not classical school, Smith and Ricardo), but by variations of income. Even if the MPC is one you still have a multiplier, since workers' income is less than total income. Kalecki's version of the Principle of Effective Demand does not require the psychological law, that the MPC is between 0 and 1. And that would hardly make the whole thing revolutionary. By the way, the important point is causality and adjusting mechanism. In the mainstream neoclassical story Investment adjusts to full employment savings, since the rate of interest does the job. In Keynes, investment determines savings, since as spending (investment) increases, income goes up and so does savings. So investment determines savings and the level might not be one at which there is full employment of capital or labor.

Keynes point was that even with price flexibility (any price, including real wages and interest rates) the system would not go to full employment. For that you must go the arguments of chapter 19 of the General Theory. So even if you have unemployment, and that would, according to Keynes, reduce the real value of money and reduce the rate of interest and stimulate investment (the Keynes' effect), he suggested that lower prices would also lead to an increase in the real value of debt, and if investors where indebted, to a collapse of investment and output. This is known as the debt-deflation effect. The discussion of the multiplier that you make is very simplistic, in all fairness. There are problems with Keynes' theory, but they actually stem from the mainstream discussion of investment that he accepts in his book. In particular, the evidence on investment is well established and is explained by something called the accelerator. Firms buy machines in order to keep capacity adjusted to growing demand, so investment is not autonomous spending, which is the central force behind the multiplier and the determination of output and employment. Also, investment is impervious to changes in the rate of interest.

I recommend you check first a good manual, or read Kalecki, which was an engineer and developed Keynes ideas at the same time independently. For the more advanced problems I suggested you may read this and the references at the end I think the key book is still Amadeo cited in my references.
Best of luck.

Matías Vernengo

You must have Javascript enabled to use this form.



Thank you for the critique. I am familiar with the neoclassical interest rate adjustment argument (which I also reject). But I also reject Keynes more fundamental argument of any kind of link between S & I. I realize Keynes was writing when the world was on a gold standard, so a comparison today may or may not be fair, but if I'm my memory does not fail me, credit was extended far beyond savings in the form of gold. I think the implications are easier to see in a modern economy. Say no one saves at all. Modern credit can still be created. Several new books on banking in modern England argue this point. Thus in my view there is no fundamental link between investment and saving. As an engineer, I see investment as a function of technology required to execute the lowest cost product and consumer demand (savings determined by consumer, not investment) for a given product. In other words, if 10% vs 25% of GDP is required given a state of technology, in both cases full employment will be achieved if consumer spend their income. Maybe I'm missing a fundamental point?

I'll check out your links you suggested.

You must have Javascript enabled to use this form.