On the surface, Keynes' theories seem to have some plausibility, mainly thanks to the simple, but seemingly logical mathematics he provides. While they are hardly as impressive as some of the byzantine calculations developed by his neo-Keynesian successors, Keynes' "proofs", from his simplistic "national income" calculations to his "multiplier" gain a lot of credibility from their mathematical nature. People simply are much more inclined to accept numbers, and, more important for Keynes, have trouble refuting an argument if the numbers add up.

But calculations only work if they are based upon valid numbers. As I discovered when studying something completely unrelated to economics¹, if you choose your measurements incorrectly (or perhaps correctly, if your intent is to deceive), you can create a situation where the calculations seem to produce one result, while a different measurement would produce entirely different results.

And in Keynes' case he did precisely that, choosing to work in both aggregate figures and monetary measurements, while arguing for inflationary measures². Actually, both are aspects of the same problem, as aggregating necessarily involves converting quantities of goods and services into their monetary values, and then adding those values together. So an aggregate value for, say, national income, is simply a monetary figure, hiding many important details, such as the relationship of various prices, the rising and declining values of various goods, the relative distortions caused by inflation and so on. I know Keynes dismisses such details as irrelevant, but in reality such changes are quite significant for his theory, he simply makes them invisible by using this purely monetary figure.

So, why are monetary figures so problematic? Don't economists always use monetary values in their calculations?

Obviously they do, or do for the most part. But in most cases, in traditional "microeconomics", those monetary figures are used for an instant of time, for a "snapshot" during which time money can be considered essentially constant in value. Also, they are used to consider the value of a single good or service, or a small group of goods and services. For example, to argue that if adding another worker would give $10 in value, while costing $9, an employer would hire that worker, or saying that with such a cost structure and such a demand curve, producers would supply X widgets. In this context, monetary values are legitimate measures, as they do not obscure important details. Keynes, on the other hand, uses money as a sort of sleight of hand. His "multiplier" for example, which argues that adding X dollars to the economy, brings Y dollars of "activity", is simply absurd, though it does show the shortcoming of figures like the GDP.

Allow me to explain.

"Activity", while much valued by Keynesians and government economists, is really meaningless. As you can see if you consider two economies. In one, two men exist. Both are self sufficient farmers, who feed themselves but trade nothing. In this economy, the GDP is $0, and there is no activity. On the other hand, let us picture a second economy, where two men both grow one apple each during the entire year. However, each buys the other's apple at $1. This system has a GDP of $2. Thus, according to Keynes and the GDP, the second economy is more healthy. In reality, as anyone could see, the second economy will result in two farmers dead of starvation before the next harvest. And it gets worse. Suppose you want to really improve your economy. You just ask the farmers to sell the same apple a few more times. If they sell the same apples again for $2, then for $3, and then for $4, you now have $20 of activity³. There has been no change in the underlying economy, but on paper there is now much more "activity". More important, not only has the GDP risen tenfold, but the assets of the nation have now gone from $2 worth of apples (that is 2) to $10 worth of apples (still 2).

And that actually points out the second problem with Keynes, and the second reason he loves monetary measures. As an advocate of profligate spending and inflationary monetary policies, Keynes is well aware that his yardstick is made of rubber, that the money he is using as a measure will be losing value over time, causing numbers to rise without any change in the underlying real value. In fact, he relies upon that. And before anyone dismisses this, Keynes himself says so in one context, arguing that monetary inflation will ameliorate the harm of union contracts, but reducing the real value of wages, making union salaries fall into line with wages in general⁴. But that is the single case in which Keynes recognizes that monetary and real values are not identical, in every other case I can recall, Keynes seems to confuse monetary value with some sort of real worth.

Lest someone think, as did many neo-Keynsians, that this defect can be remedied with some sort of "deflator" figure, there are three problems. First, any deflator is itself an aggregate figure, and inflation does not strike uniformly. If we use, say, the CPI, we are seeing only the inflation in the consumer sector. But in our most recent inflationary period, inflation struck the housing sector first, driving up the cost of real estate, and then construction materials, which means a CPI-based deflator applied to the economy as a whole would insufficiently deflate. On the other hand, if prices rose first in the consumer sector, the CPI-based deflator would deflate the economy too much. Second, there are price changes, even during inflation, completely unrelated to inflation. For example, energy prices in the 70's rose for reasons unrelated to monetary expansion⁵. We can try to account for this, but often trying to guess which prices are rising at any given time for "legitimate" reasons and which for inflationary, is a very speculative exercise, and makes the deflator less and less objective, leaving lots of room for personal opinion to masquerade as fact.

Then there is the final problem, and a problem which plagues Keynes' theory as a whole. Inflation does not just change prices. If that was all it did it would not be half as harmful. But inflation also changes the behavior of individuals, causing the economy to distort in unexpected ways. And many of these will cause changes in economic "activity", though without really changing the underlying value of the economy. For example, early in an inflationary cycle, when there is additional money available, but people have not yet come to fear continuing inflation, borrowing may increase. This could lead to people investing more, as demand for money races ahead of the extra money available. Or it may cause people to borrow, as interest may be low. In either case, it changes the aggregate figures about which Keynesians obsess, making the economy appear more "active" and hence healthy, while in reality the economy is actually becoming ever more distorted and unstable. Even worse, as the economy begin to falter, once people come to realize inflation will continue, that their money will regularly lose value, individuals will try to exchange any money they hold for goods as quickly as possible. Obviously someone will have to hold the money, as, despite Keynesian claims, money does not "circulate", at any instant someone is holding it, but they will each try to hold as little money as necessary for as short a time as possible. This will result in a huge "multiplier" effect. In Keynesian terms, this is a good thing⁶, allowing very little activity to boost the economy. And there really cannot be much greater an indictment fo Keynes. That his theory confuses the frenzied buying of incipient hyperinflation for robust economic growth is an indicator of just how far removed from reality Keynesianism is.

The truth is, money is a good. When money was simply a representation of a commodity, like gold or silver, this was much more obvious. In the 19th century, for example, those int he US pushing for "easy money" to allow creditors to pay off debts in deflated currency, tended to push for defining the dollar in terms of silver rather than gold, as recent discoveries of silver deposits meant silver was dropping in value, making it effectively an inflationary currency, while the "hard money" side pushed to retain the gold standard, as its value had been more consistent through the century. But in both cases, those making the arguments were aware money was a commodity, rising and falling in "price", that is the rate at which it exchanges against other goods⁷.

Keynes seems to either forget, or conceal, that basic truth. He treats dollars as if they were inches or grams, something with a fixed, immutable definition, and then proceeds to use them in calculations, completely ignoring the fact that his "inches" can change in the course of an equation. And just as construction would be difficult if a foot sometimes had 12 inches and sometimes 18, Keynes' calculations prove rather misleading when dollars can change in value. Though I tend to think this is intentional, as the "mistakes" tend to point all in one direction, toward ever more government spending and monetary inflation, I suppose it is possible Keynes deceived himself as much as he did the rest of us. Then again, that is really irrelevant, what matters is that he was wrong, and many of the phenomena he observed were mere optical illusions, caused by the shifting values of currency.

I could probably go on at much greater length (as any regular reader can attest), but I think this more than makes my point. Monetary calculations are fine, without them economics would be much more confusing, but we also need to bear in mind that money can change in value, and that monetary measures do not have the precision of weights or lengths. Much of what we see in the economy can be reliably measured in currency, but we need to be careful. Sometimes we can mistake monetary phenomena for economics ones, and that can lead us into very serious trouble⁸.

-----------------------------------------------------------------------------------------

1. Mainly this came up when I was asking about the "Twins Paradox". It just seemed to me the argument always smuggled in an assumed fixed frame of reference, while the whole thesis was that all motion is relative. If you can view one twin as accelerating and moving, then decelerating and returning, you could just as easily view him as fixed and the other twin doing the same, so it seemed both should be older than the other. Which led me to wonder if the problem of the theory may not be the way Einstein chose to define time. I never really found an adequate answer, nor decided if my concerns were legitimate, but it did turn out to have interesting implications for other areas of study. (I know some experimental and observational evidence supports the theses about time, such as the decay rate of high speed particles, so I am inclined to think my problem is so far getting bad explanations, but it is still possible there is some other explanation, not involving time, or, perhaps, there is a fixed frame of reference despite the basis of the theory. I know that is heretical to some, but it is still a possibility. And theories do change. After all, Einstein himself couldn't decide if foreshortening was real or apparent.)

2. He also proposed some self-contradictory theories. For example, after saying savings and investment  must always be equal, he premised his entire theory on imbalances between savings and investment. (Which, excluding inflation or an absurdly high preference for cash holdings, is an impossible situation.) Even though his conclusions would still be wrong, even if savings and investment could deviate from one another as much as he suggests, the fact that the two numbers really do follow one another closely suggests Keynes is fighting straw men of his won creation.

3. When real estate investors sold the same houses back and forth to inflate the value, allowing them to support large loans, it was called "flipping" and was illegal. (And rightly so, as the borrowers often then defaulted on those loans, leaving the lenders stuck with overpriced foreclosures.) When the government suggests doing the same tot he economy as a whole, it is called a sound economic development plan.

4. Amusingly, the one place Keynes recognizes real prices differ from monetary prices, he is wrong in his prognostication. Unions, being well aware of the effects of inflation, began adding COLA escalator clauses to union contracts even before Keynesian policies were consistently implemented, making the single semi-plausible economic benefit of inflation disappear.

5. Which is why many deflators started to use CPI with energy removed. But, as energy costs are part of almost every consumer good, whether form production or transportation, it is almost impossible to truly remove the effects of energy from the CPI. So every deflator from the 70's (and from the mid-2000's as well) is somewhat tainted by incorporating price growth unrelated to monetary expansion.

6. Keynes does make some absurd statement about when the multiplier grows too great prices can rise "without limit". But this is quickly ignored, and Keynes returns to his general praise of the beneficial effects of government "investment" due to the multiplier effect. So, yes, he does admit at some point his "multiplier" can be harmful. But as his "multiplier" is wholly imaginary, and his love of "circulating" money is misdirected, that really isn't something to praise.

7. This is a problem in economics in general. In "microeconomics" we tend to define "price" as the amount of money required to purchase a quantity of goods. However, this definition tends to assume money will remain fixed in value. (As expressed in the "cet. par." following every equation.) So we assume any changes in price are the result of shifts in supply and demand relating to the good. On the other hand, when we shift to "macroeconomics", we tend to assume supply and demand remain fixed, and any change in price is a wholly monetary phenomenon, related only to shifts in the supply and demand for currency. In the real world, we often can't tell with certainty what is causing a shift in prices, even one effecting many goods. As I mentioned in footnote 5, when oil prices rose in the late 70's, during the course of Carter's inflation, it was impossible to tell how much of rising consumer prices was due to increased cost of energy and how much was due to inflation. As both caused price changes at every point in the manufacturing, distribution and sales process, it would have been impossible to assign a percentage to each with anything approaching certainty.

8. To provide just one example, it is unthinkable that a pre-Keynesian government would have decided the best solution to a deflationary crisis would be spending nearly $800 billion on random government expenditures. Prior to Keynes, the normal response to slowed growth and reduced tax revenues was less spending. It was only after Keynes that "spending" became "investment", and the idea spread that government profligacy was the path to riches.

----------------------------------------------------

POSTSCRIPT

If anyone is interested in my other writings on Keynesianism, and inflationary monetary policy in general, I suggest examining the list of articles included in the postscript to "Living Large During the Good Times". It is as close to comprehensive as I can come.