Solow vs. Solow

Robert Solow writes,

The custom is to think of value added in a corporation (or in the economy as a whole) as just the sum of the return to labor and the return to capital. But that is not quite right. There is a third component which I will call “monopoly rent” or, better still, just “rent.” It is not a return earned by capital or labor, but rather a return to the special position of the firm. . .The division of rent among the stakeholders of a firm is something to be bargained over, formally or informally.

Pointer from Mark Thoma. If this is true, and I believe it is, then the marginal product of a worker cannot be reliably measured, nor can it be proxied by a measure of the wage rate. That makes some of Solow’s early empirical work, and much that followed in its wake, somewhat problematic.

Recall what I wrote.

Separately, each worker’s contribution to the process is not marketable. It is the final product that can be sold. In a sense, there is a “production externality,” in that the finished product is worth something, even though the individual worker’s output is worth nothing by itself. The task of Coasian bargaining among the workers to come up with a way to allocate this externality is onerous, so it is handled by a manager in the context of a firm.

Also, see my post on the Alchian-Demsetz theory of the firm.

17 thoughts on “Solow vs. Solow

  1. “Separately, each worker’s contribution to the process is not marketable. ”

    That doesn’t seem right. Consider an accountant working for a widget manufacturer. On the one hand, of course the firm sells widgets and not quarterly statements. On the other hand, the accountant’s services certainly are marketable and his pay is determined primarily by the market value for accounting services rather than via either ‘Coasian bargaining’ among co-workers or by a manager (who, after all, cannot dictate the market value of accountants).

    Or am I missing something?

    • The accountant has opportunity costs, which puts a floor under what the firm must pay to get the accountant to work. But that does not measure the value of the accountant’s work inside the firm.

      • This starting to feel a bit esoteric. Consider the steel that goes into making those widgets. Say it costs $500/ton on the open market. Does it make any sense to say that the value of steel ‘within the firm’ is something other than $500/ton? How are accounting services (which might be internal or outsourced) different, in principal, than steel?

        • Yes, I think it does.

          Change it to mouse traps and spring steel. What if spring steel prices go up to the point that mouse traps aren’t profitable and the company has to go bankrupt?

          • (if it wasn’t hard to think about this army of Nobel prize economists would have already figured it out 😉

          • What if the steel drops to $250/ton — is the ‘intrinsic value’ within firm still $500? Is the intrinsic value defined by ‘the maximum price the firm could pay for the input without going under?’ But that doesn’t work — since that amount depends on what is being paid for other inputs (if labor prices double then even cheap steel might put the firm under — but was it the cost of steel or labor that was ‘responsible’?)

    • Another angle is that even though accounting is pretty standardized, what the accountant does for his particular firm is not fully commodotized accounting.

      In the education discussions I like to point out that I solved hundreds of distillation columns in education and will never see one, lord willing, in my career. Every job I’ve had involved very idiosyncratic tasks unique to not only that type of job, but that particular firm.

      There is a labor market for people like myself, or so I hear, but not really a market for what I had to learn to do that added value at my actual jobs. Indeed, that human capital that added value to me within my particular firm, kind of subtracted value from my human capital in the labor market.

  2. I like it because I’ve always bristled at the fact that the more I help the firm the further I fall behind. How to make individual skills more marketable?

  3. Why has it become so difficult to recognize that “incomes” (wages, etc.) are only “one” (in various forms) of the allocations of the results from production, distribution, and advancements (quality, variety, innovation, burden relief, etc., etc.) of our form of social organization.

    There has been some attention given to those allocations (services and money) now channeled through governmental or politically directed activities. An extreme example might be EITC, coupled with Food Stamps, plus allocated housing (subsidized- others bearing part of the cost). All that has to come out of the organizations production. The allocation does not occur by allocating “income” but through alternative “distribution” ( often labeled [mislabeled?] redistribution). More than 60 % of the tax reporting population is now within the network of that alternative system of allocation.

    This is no longer a purely “income” or “wage” economy.

    We might consider why that is; how it comes to be; how it is evolving – and critically, where will it lead in changes to the social organization.

    • First thought is that dealing only in aggregate economics would make that investigation difficult.

      • @Andrew;

        If that was directed to my question, I would suggest that the inquiries would certainly extend into the factors shaping changes in the social organization, not be limited to just “economics.”

          • @Andrew

            From what I read (rather eclectic) there is not **an** inquiry. But there are studies of different factors and how they “might” affect the social order (including relation to changes). If there is an effort to assemble the information and perceptions for purposes of finding links in the bits of information studied, I am not aware of it.

            The nearest analogy would be the studies to determine “causative” factors for historical changes in social organizations (McCloskey, or Acemoglu, e.g.).

            There remain the academic tendencies to examine what “ought” to be causative factors (confirmation bias).

  4. My idea would be to give every employee a minimum wage from the CEO on down and then give each employee a share of any profits. The number of shares an employee would get would depend on the job of the employee. Sort of a prirates of the Caribbean approach but with profit not plunder.

  5. I’ve always thought that “workers are paid their marginal product” was a misleading way to articulate the neoclassical theory of labor markets, for two reasons.

    First, workers might not have a well-defined marginal product, as with Leontief technologies where different inputs must be combined in fixed proportion (since then the left and right derivatives do not equal).

    Second, “marginal product” only gives you one quantity-price relationship – effectively, it’s an indirect way of talking about the labor demand curve, where given a wage cost of labor you’ll hire up to the point where the marginal worker’s productivity equals the cost. The supply curve, equally (or more) important in determining equilibrium, is missing.

    The more robust and insightful way to describe neoclassical labor economics (and even some extensions on it) is simply to say that it’s about *supply and demand*. This is much less contingent on idiosyncrasies like the differentiability of the individual firm’s production function.

    This broad understanding of neoclassical economics can easily handle cases, by the way, that you view as requiring some more exotic model. Suppose that each firm operates a production technology that uses some fixed-coefficient mix of Economists and Lawyers to produce Widgets. In none of these firms, individually, is there a well-defined marginal product of Economists or Lawyers. But as long as there is (smooth) heterogeneity among firms in their input coefficients, the forces of supply and demand still pin down the relative pay of Economists and Lawyers, and these will reflect their *social* marginal products.

    (The idea that fixed coefficients for individual production units can combine with heterogeneity and the extensive margin to produce elastic aggregate responses is not new; it was the premise of Houthakker’s classic paper showing that Cobb-Douglas aggregate production could result from Pareto-distributed input coefficients.)

    Indeed, you don’t even need this aggregate smoothness on the demand side as long as there’s smoothness in substitution on the supply side – and surely that exists, since each additional increment in relative wages will prompt some Economists to become Lawyers, or vice versa.

    In general, I suspect that long-run elasticities of substitution are higher on the labor supply than the labor demand side, at least in a lot of cases. (The technological ability to replace one kind of labor input in a manufacturing process with another may be nearly zero; but individual workers’ willingness to provide one kind of input relative to the other may vary tremendously as the relative wages change, especially in the long run as they can train accordingly.)

    This high elasticity means that even seemingly large shocks to the labor demand curve – which might imply big shifts, holding the quantity of labor input constant, in relative marginal products – may have a much smaller *equilibrium* effect on relative wages than one might assume. All the more reason why “marginal product” is a poor way to summarize the neoclassical view on labor markets…

    (Addendum: supply and demand also encompasses the possibility that firms make *mistakes* in assessing the importance of their inputs. Then the inputs aren’t paid their true marginal products; yet supply and demand still matter, and the welfare implications of firms’ distorted judgment depend on supply and demand elasticities, etc.)

  6. I don’t know, this all seems very Wittgenstein-ish to me. It’s just argument by label: It is a rent, It is not a rent; it is too a rent, etc.. Mary labels things one way and Molly another. None of it is empirically demonstrable. And if you believe it then nothing about productivity is empirically provable.

    The other thought that occurs to me whenever I read these kinds of articles is that the data is just not complete, in particular that the “wages” are not being properly measured, because the data often excludes health insurance, pension accruals and other costs of employment, and often exclude supervisory workers’ compensation. So I don’t know how anyone can extract an inference about the data with such incompleteness. It’s possible that supervisors are more responsible for productivity than is being recognized.

Comments are closed.