Money is not an epiphenomenon: the unreality of the “real”

There is no such thing as “real wages”. Economists talk about “real wages” but what do they mean by that? In a monetized economy, prices and costs are measured in money terms. So, a firm has to worry about the terms of labour—the ratio of labour costs to the prices of what it sells (weighted by how much labour produces how much product: so constraints on the use of labour will still affect the terms of labour). It experiences these things, and makes judgements accordingly. If the ratio moves adversely (which might be because the prices of its products have fallen, or because they have risen less than labour costs: any movement in either labour costs or output prices is potentially compatible with an adverse shift in its terms of labour, since it is the ratio of the two that matters) the firm will tend to cut back on hiring. If the ratio moves positively, the firm will tend to increase hiring.

But the terms of labour are not the same was the terms of wages—the ratio of payment to the worker to the prices of what the worker purchases. First, because wages received are not the only labour costs. Second, because what a worker purchases has no particular connection to what the firm sells. Indeed, different workers will have different terms of wages, since they will not have exactly the same pattern of purchases; just as the terms of labour will vary between firms and, even more, between industries.

So, when economists talk of “real wages”, what do they mean? Do they mean the terms of labour or the terms of wages? If they mean both, they effectively mean neither but instead some mystical (because very unclear) amalgam of both which is not specifically either.

And how do we measure “real wages”? By “deflating” wages (which are not the price of labour, but leave that aside) in terms of some general price index? If it is some specific index, such as the CPI, then that is yet another general amalgam which, at best, crudely correlates to what either the worker or the firm is experiencing and making their judgements about. If it is a general measure, such as a GDP deflator, it has less arbitrary selection problems but still has only a crude connection to what the worker or the firm experiences and makes judgements about.

But, why bother going to that effort? The process of deflating adds nothing to the information to be had from terms of labour and terms of wages. Indeed, it is worse than that, because it actually takes information away. Sticking with the money prices and costs both reflects what people actually make judgements about and does not lose information.

The underlying idea that the notion of “real wages” taps into is that money is some epiphenomenon which overlays a “real” economy. But, as we have seen, trying to postulate something called “real wages” fails to pick out a specific phenomenon, suppresses information as it does so and does not capture what people actually make judgements about. If our concern is with human behaviour then the issue becomes what information do people use to make their decisions. The notion of “real” wages fails to accurately capture any specific thing.

Consider the asymmetry between increases and cuts in (money) wages. If money was an epiphenomenon over some “real” economy, there should not be any difference between raising money wages when the price level is rising and cutting money wages when the price level is falling. But contracts, debts and financial obligations are set in money terms and operate across time periods, so there is a clear difference between the two. Cutting money wages increases the burden of existing debts and obligations. So, it is perfectly rational for workers to resist cuts in money wages even if their general terms of wages are rising. Looking at “real wages” again suppresses information; indeed, it seriously misleads.

Nor is the notion of “real prices” any better. There are only two sorts of prices: money prices and barter prices—prices in terms of money and prices in terms of other goods and services. The first can be expressed in a common range of numerical values, a measure of prices that operates across goods, services and assets. It is one of the great advantages of money. The second can only be expressed in terms of other goods and services. The notion of “constant price” is not a “real” price: it is simply prices expressed in “frozen” money abstracting away from general shifts in money prices/the barter prices of money. One is using a key characteristic of money while pretending to get “past” it. To so attempt to use money to get “underneath” to the “real” economy is to, in fact, express how much money is not an epiphenomenon. One is still using money, just in a particular way.

Yes, people are aware of shifts in the barter prices of money: which is to say, the inverse of money prices. But there is not some “real price” beyond that.

Money is a transaction good: people use it to transact to get the goods and services they want. So, if we have three goods in an economy (consumption, assets and money) then we have two markets (money for consumption goods, money for assets). The process of transacting uses a medium of account (money) and does so for good reasons. The advantages of money over barter are not some epiphenomenon. They are major advantages that change how people behave and so how the economy works, particularly given money operates across time periods (we can spend now or later; we have previous entered into obligations expressed in money terms).

Using the concept of "real" prices abstracts away from “actual” money while continuing to invoke its functions. This is not analytically helpful, for we then make money what it is not—immediately, transparently “neutral” about prices in terms of goods and services. Cognitive simplification—being able to express prices in common numerical values—means precisely that and is a genuine economic function. It takes time to register general shifts in the barter price(s) of money and for credit, contracts and other prices to adjust. Which means that shifts in spending have effects on output, until people adjust for any general change in what money buys (in terms of goods, services and assets). By "abstracting away” from money (even though we are actually not fully doing so) we also abstract away from money being a cross-temporal constraint due to contracts, debts and other financial obligations.

The notions of “real wages” and “real prices” do not get to “underlying” realities, they obscure economic realities because they abstract away from how people are actually making decisions and constraints on those decisions. Money is the prime form of information in a monetised economy. By treating it as some epiphenomena, we are not revealing, we are obscuring.

Money is not an epiphenomenon and it is actively misleading to use economic language that implies it is: particularly when such language ends up suppressing relevant information.