A sideways look at economics
Groucho Marx once stated indignantly, in a beautiful paraphrasing of Russell’s Paradox, that:
“I wouldn’t be a member of any club that would have me as a member.”
It’s important not to confuse Groucho Marx with Karl Marx, the double centenary of whose birth is this year. They share little in common apart from the name and unconvincing facial hair. But there’s a nice application of Russell’s Paradox to what remains of Karl’s legacy.
Marxism is bunk. The doctrine of the historical inevitability of revolution is superstitious nonsense, akin to believing in the tooth fairy, and has been comprehensively dispatched by Karl Popper in The Open Society and Its Enemies. Karl Marx’s Das Kapital, apart from being almost unreadable in its tedium, is littered with pseudo-science (a charge that could be fairly laid at the door of much of what passes for mainstream economics too). And so on.
Of all the bunk, the most bunky (henceforth a word) part of Marxism is generally held to be the labour theory of value (LTV). This theory posits that the value of something is determined by the quantity of labour that has been involved in producing it.
Marx applied the LTV to what he called ‘commodities’ (meaning items that are manufactured and are identical in all important respects). Marx acknowledged that the price achieved by such items in a capitalist economy is not determined by the quantity of labour input, but by the interaction between supply and demand. But he asserted that the value of a good is different from its price. (Economists, as the cliché goes, understand the price of everything but the value of nothing — this is ultimately a Marxist sentiment.) Capitalism, he argued, drives down the cost of labour to the bare minimum consistent with the survival of the worker, while the difference between the market price achieved for the goods those workers produce and the ultra-low cost of labour is the ‘surplus value’ that is ‘expropriated’ by the owners of capital from those to whom it rightfully belongs: the workers. Socialism would, by taking control of the means of production (essentially, seizing the fixed capital assets in the business), distribute that surplus value to the workers rather than to the owners of capital. And, in that world, the price of a good would equal its true value which would be proportional to the quantity of labour input. That quantity is the heart of the concept of value.
The received wisdom on this point is that it is the utterest (also, now, a word) of utter tosh, since value is in the eye of the consumer, rather than of the producer. Imagine, for example, if someone spent all day every day knitting scarves that nobody will ever even look at, never mind wear. The value of that activity — and of the scarves — is zero, beyond any pleasure that the knitter derives from it, irrespective of how much labour is employed.
What’s peculiar — paradoxical — is that Marx’s ‘commodities’ are the very things to which the LTV is least applicable. Generic, non-branded (or let’s allow weakly branded) manufactured goods — think of a clothes rail, a peg, a car mat, a bin liner — have a clear value for the consumer, which is largely unrelated to the quantity of labour employed in their production. The market will clear at the competitive price irrespective of the technological mix used to produce the goods — the factory could employ one person to monitor the machinery, or it could employ 1,000 people. It’s irrelevant to the price, and therefore to the value.
On the spectrum of relevance of this strand of Marxism, commodities mark zero. You wouldn’t let Marx into that club, so to speak.
But at the other end of the spectrum are services like those that are provided by Fathom Consulting. Marx might well be considered as a member of this club — but I strongly suspect that he wouldn’t want to join.
There’s a general and difficult question about how the value of services should be measured, and this applies particularly to consultancy and other professional services. When we bid for a contract, we’re nearly always asked to provide a detailed breakdown of the number of person-hours that we will employ in delivering it. And, of course, lawyers and accountants are well known for their ‘billable hours’. Why do the hours matter? The key concept ought to be the additional value that we provide to the client. The hours we take to create that value should be as irrelevant as they are in markets for Marx’s commodities. The problem is that the client for consultancy (or other professional) services is rarely able to estimate accurately the additional value to their business (or government department) that accrues as a consequence of engaging our services, and is even more rarely able to distinguish between the value that accrues from engaging consultant A compared to consultant B. These things are hard to be sure about, especially ahead of time.
From our side of the table, it’s equally hard to know ahead of time how many hours we’ll expend on each contract. What proportion of the hours we spend generating research for all our clients could or should be attributable to this particular contract? What about the hours spent thinking about this contract on the train, or at home — do they count? What about the time we spent talking around it in the pub (don’t worry, only if it’s not sensitive) the other night — does that count? How many hours did I work last week? How many hours did other Fathom staff work? How were they split? What about the hours spent looking at Facebook while simultaneously mulling over how to articulate the econometric model that will be the key to solving this client’s problem? Etc.
The consequence of this uncertainty is that when we bid for projects, we state what we believe to be the expected effort deployed in delivering it, but we acknowledge to ourselves that the risks around that expectation are large. Because our reputation is the core of our brand, we can never afford to let our clients down, so we will put in whatever effort is needed to see the project to a satisfactory conclusion. And if we run over in terms of hours, we’re almost never able to pass those additional costs on to the client, but instead have to wear them in reduced profits/salaries/bonuses.
Now, if we routinely under-estimate the effort involved, then we‘ll go out of business. And if we routinely over-estimate, then our clients will be able to find other contractors who would do as good a job for less money. And that logic is writ large across our industry.
Moreover, the hourly rates we can charge are fixed within fairly narrow ranges, that once again apply more or less across the whole sector. The client will typically be aware of those ranges, and if they don’t know much about us then we either have to be able to make profits at the market-implied hourly rate or, ultimately, go out of business.
The market for consultancy services clears at a price that values that service based loosely on the number of hours employed, for something close to a fixed hourly rate. All of which means that the LTV applies, more or less, in our industry.
So, the club that Marx wanted to join (as it were) — manufactured goods — wouldn’t have him as a member. But the club that would let him in — professional services — he wouldn’t consider joining. Bertrand Russell would be proud.
In other respects, our sector is nothing like the Marxist model. We’re not wage slaves but are paid reasonably well. We’re not alienated from our work but deeply engaged in it. And nobody expropriates the ‘surplus value’ we create except, perhaps, the government — though our clients share it with us. We are for profit: strongly so.
If consultants in general could all agree to increase hourly rates, that would show up, in aggregate, as in increase in productivity in our sector — even though it wouldn’t necessarily imply a general increase in quality. With no change in average quality, a higher hourly rate would simply mean transferring resources from our clients to ourselves, and it’s sensible to assume that aggregate demand for our services would fall. The Marxist exhortation doesn’t work for us: consultants of the world, unite! We have nothing to lose but our clients…
 Which asks the question whether the set of sets that are not members of themselves is a member of itself. If yes, then no; if no, then yes.
 There’s a vast literature on this topic, and it’s obviously impossible to do justice to it here. This passage is shorthand and will doubtless strike many as grossly distortionary. Sorry about that.
 Unkind readers might argue that’s because the value is insignificantly different from zero, but there’s just no need for that kind of negativity here.
 Difficult though this is to imagine.