A very interesting point in macroeconomics is made by Scott Sumner (http://www.themoneyillusion.com/?p=8360) regarding the possibility and plausibility of zero marginal productivity workers. Upon observations that
The ZMP discussion seems to me to flirt with the edges of the fallacy of misplaced concreteness (N. Georgescu-Roegen, A. N. Whitehead). MPL is not a characterization of the last employee hired *only*, in part because the capital employed forms a constaint and people “step on each other’s toes.” (see, e.g., http://en.wikipedia.org/wiki/Marginal_product_of_labor#cite_note-Perloff176-5). This is indicated explicitly by Samuelson in his economics textbook, *Foundations of Economic Analysis*, Harvard, 1947:
“[I]t is important to underline that the marginal product is not, properly speaking, the econtribution of the marginal unit *by itself*. Some commentators (e.g, E. v. Bohm-Bawerk) seem to have gone on to make arguments that seem to imply [that the new worker added to the enterprise produces exactly the amount reflected in the marginal product]. Of course this is not necessarily true. The second man may very well produce more or less and still the total output increases [because the second man reduces his output correspondingly]. [B]y *adding* the second man, output was increased by [say] *eight*. *Thus*, the marginal product of the second man is eight. But his actIual contribution may be very different than this.” [all emph. in original]
Therefore to say that *the worker has a MP of zero* seems inaccurate. Labor for the economic process taking place in a particular enterprise exhibits MP of zero, not the laborer. *It is a characteristic of the production function*, not of the worker. The question “whether those people produced anything useful” seems therefore, in the Neoclassical paradigm, without import.
Let’s assume “we are indeed producing as much as before” and that the metric for this is reliable, apples-to-apples, etc. The workers that are let go may have brought about the situation that the firm’s production function exhibit ZMP in the industry in which they worked (because of the constraints there). It only seems that, yes, the marginal/last/final hire will have the least productivity that would “cause” the total output increase to become zero, but it does not follow that he would *possess* the least skills or is loafing—this is not just a skill problem. MPL is specific to the industry/firm/economy just as well as to the skill level (at least in the short run). Furthermore, ask any HR department and it will become apparent very quickly that the people usually let go are not the ones with least skills—last in first out regardless of skill, disguised as most-recent-non-performance, wins out most of the time.
Marginal Revenue Product of Labor is MRPL = MR x MPL, that is Marginal Revenue x Marginal Product of Labor
With the output measured in dollars (appropriately or not wrt apples + oranges), as long as MR is not zero, the GDP measures are connected to MRPL, rather than MPL. And MR can be greater than or less than 0, depending on the price (in)elasticity of (aggregated) demand. So there may be an ambiguous signal here—people were let go because the marginal revenue on their output went down OR the marginal productivity of labor did. The “recovery” may have “recovered” MR (e.g., made it >0 or even =0(!), the specific case of contention) *at existing employment level’s MPL*.
People become analogous to the horses [another commenter] mentions (http://blogs.law.harvard.edu/philg/2010/08/08/unemployed-21st-century-draft-horse/) *only if* the equilibrium wage, w = MPL < s, where s is the subsistence wage. Note that this may happen even if the demand curve for labor (determined by MPL) crosses the labor supply curve *but below s*–the income allocation issue becomes thorny and “Malthusian” at that point *if* the allocation method also tries to use the assumption of strict hedonism and profit maximization (w=MPL)—a characteristic of an (non-strictly) overpopulated labor market. Moreover, it may be the case that MPL hits the “x” axis without ever crossing the labor supply curve—a strictly overpopulated labor market. (apud N. Georgescu-Roegen, “Economic Theory and Agrarian Economics”)
Scott clarified the point that “if firms are fairly monopolistically competitive, then the MR may be fairly low at current output. Combine that with a fairly low MPL relative to the APL, and you’ve got a really low MRPL. Maybe not zero, but pretty close. That would be Tyler’s best argument.” Makes sense.