I was reading the last few chapters of Keynes’ General Theory, and the discussion of various under-consumption theories prompts me to discuss the most important critique of this approach. Further, this critique also serves as a powerful reason to doubt the logic of the Keynes–Kahn multiplier, where the inducement to invest is a direct function of the marginal propensity to consume. The argument follows from Böhm-Bawerkian capital theory, and is that a large fraction of investment is meant to satisfy consumption at some point in the future. Production is a process that takes time. The current rate of consumption has little to do with the current volume of investment (although, as I’ll argue after, this doesn’t necessarily imply that the stock of capital is what decides the volume of investment).
Keynes is correct as far as the proximate cause of investment goes, and we’ll use this as the starting point of our analysis. An entrepreneur will invest when expected revenue is equal to or greater than her costs, which Keynes defines as factor costs, which is what firms spend on inputs, and interest. But, and parting with Keynes here, expected revenue is a function of future consumption (or, what we expect future consumption to be), not of present consumption. Further, allowing for the imperfections of reality, the factor costs will also be decided by future consumption. We know that producers’ good prices are imputed from demand for final goods, and that this is achieved through competitive bidding of property, but the value scales of sellers and buyers will reflect expectations of the future. In other words, the present prices of the factors of production will be decided by the future output they’re going towards the production of. Thus, as the fraction of income consumed falls so will present input prices, and total income will be spread over a longer structure of production.
Another point is that current income in the capital structure is paid out of savings, not consumption. Wages, factor prices, et cetera are all paid out of savings. Keynes, Kahn, and many others, got the process exactly backward. The consumer doesn’t buy a product, which then stimulates that firm to buy inputs for output, and so on and so forth. Rather, a firm, at some point in the past, believed that at some point in the future its product would be demanded. To pay for the production process a firm has to dip into savings, whether out of its own retained earnings, or by borrowing from a bank, or even from income earned from interest bearing assets (e.g. wholesale credit markets).
The point may be clearer if we look at the problem as it would exist in a simplified world. Suppose that at point t the stock of all goods is equal to x, and that x – y is consumed (y ≡ capital stock), which we’ll call z. In order to produce z consumers’ goods, you need a capital stock of y. At t + 1, society’s time preference falls such that a stock < z is consumed, in return for a future stock of consumers’ goods > z. To produce more output, though, you need more inputs, and only savings make these inputs available. In The General Theory, and elsewhere, the logic is opposite: an increase in savings ought to follow an increase in output (in Keynes’ terms, first investment rises and then in the next period the rate of savings equalizes the rate of investment). But, in my opinion, this way of seeing things is completely erroneous.
As I provocatively suggested in the first paragraph, none of this means that the rate of investment necessarily follows from saving. Keynes’ under-investment argument has to be argued against from a different angle. Even without the multiplier, Keynes’ case mainly hinges on the possibility of the rate of interest being above the natural rate, such that the marginal efficiency of capital is lower than the costs of borrowing money. This would occur if the market rate is also determined by liquidity preference, where a rise in liquidity preference means that a price has to be paid such that people will part with liquidity. This requires a completely different set of counterarguments (some include the flexibility of the stock of inside money in a free banking regime and means of reducing the liquidity premium, through securitization, et cetera).
But, the multiplier theory is no good. It depends on a relationship between consumption and investment that is the exact opposite of reality. It requires that the paradox of thrift be true, and it should be clear that it isn’t.