≡ Menu

Bryan Caplan and the Zero-Marginal-Productivity Hypothesis

As is typical of Bryan, he concisely and clearly makes a powerful argument – this time against the proposition that in today’s first-world economies many low-skilled workers have zero (or near-zero) marginal productivity.

I agree with Bryan.  This zero-marginal-productivity (“ZMP”) claim strikes me as way-strange, not least because (to put my own spin on the matter) it seems to sit poorly with the principle of comparative advantage.

So I take the liberty here to riff a bit, with thoughts on this matter that are still admittedly not yet fully – or perhaps even adequately – formed.


Even the lowest-skilled worker is capable of producing something of value.  And the value of using that low-skilled worker to produce that something, rather than using some other means  (i.e., a higher-skilled worker or a machine) to produce that something, rises the greater is the comparative advantage of those other means at producing something else.

Suppose that yesterday producing 1 fish cost Suzie 1/2 banana and cost Billy 1 banana, and producing 1 banana cost Suzie 2 fish and cost Billy 1 fish.  Billy has a comparative advantage over Suzie at producing bananas (and Suzie over Billy at fishing).  Now suppose that today – after earning her PhD in fishing – Suzie’s fishing productivity rises significantly so that now each banana, were she to produce one, would cost her 100 fish.  (Suzie is now so good at producing fish – she can produce so many of them (say) per hour – that for her now to spend any time producing bananas would oblige her to sacrifice a much larger quantity of fish than she would have sacrificed before earning her doctorate in fishing.)

Even though Joe experiences no increase in his own productivity as a banana-grower (or, for that matter, as a fisherman), his comparative advantage over Suzie at producing bananas skyrockets.

Yesterday Joe could produce a banana at 1/2 of Suzie’s cost of producing a banana; today Joe can produce a banana at 1/100th of Suzie’s cost of producing a banana.  The dramatic decrease in Joe’s cost of producing bananas relative to Suzie’s cost of producing bananas is the exclusive result of the dramatic increase in Suzie’s productivity at fishing.

At least in principle, then, the maximum number of fish that Dr. Suzie is willing to offer to Joe today (100) for each of his bananas is larger than it was yesterday (2).

In short, Joe has become comparatively more productive at bananaering as a result of Suzie becoming more productive at fishing.

Doesn’t this fact suggest that, at least potentially, the value of Joe’s marginal product rises as a result of Suzie adding greatly to her human capital?


Is it really plausible that we are today producing – or that we will soon produce – machines in such quantities and with such a range of capacities that there is literally nothing that a low-skilled worker can produce of value for other human beings that cannot be supplied by such machines at a lower cost to these other human beings?