In my economics class, we learned that in the short run, there are three stages of marginal returns: increasing, decreasing, and negative.

As a firm adds the first few units of labor, specialization allows for increasing marginal returns (Stage 1). This makes sense to me.

As one adds more units of labor, marginal returns decrease due to limited capital (Stage 2). This also makes sense to me.

However, in Stage 3, there are negative marginal returns (total product decreases). How is this possible? In practice, the firm would lay off the final worker(s), and the firm would never reach Stage 3 unless management was totally incompetent. However, by the theory of specialization that causes Stage 1, if, say, a factory could produce 30 items with 6 workers but could only produce 27 items with 7 workers, it seems as though the factory, assuming it kept 7 workers on its payroll, would simply have one worker "specialize" in sitting outside, letting the other 6 workers operate independently of the one sitting outside and thus produce 30 items. By that argument, if the firm kept hiring more workers, it could have those newly hired workers sit outside, so that total product would remain constant and marginal product would never be negative. Where am I going wrong?


  • $\begingroup$ as Alecos mentionned, a "rational" firm will produce until the marginal return reaches zero. $\endgroup$ Mar 7, 2017 at 22:08

1 Answer 1


I don't understand your problem: negative marginals returns are a feasible situation. The fact that a rational optimizing firm won't let itself reach such a stage is another matter.

  • $\begingroup$ The problem is, how can marginal returns feasibly become negative when the model for marginal returns allows for specialization, so the firm, when hiring an additional worker, could have the additional worker "specialize" in sitting outside and not interfering with the production process while letting the original workers keep doing what they were doing before, achieving a worst-case marginal return of zero. $\endgroup$ Mar 7, 2017 at 21:02
  • 1
    $\begingroup$ @GeorgeBentley Again you insert a purposeful behavior from the part of the firm. It is as though you ask "how is it feasible to die if you fall off a skyscraper, since you will never rationally do so?". The fact remains that if one jumps, one will die. $\endgroup$ Mar 7, 2017 at 21:05
  • 1
    $\begingroup$ But isn't total product the optimal number of items the factory could produce with a given number of workers? And if you are having negative marginal returns, you'd produce your optimal number of items by having some of the workers sit outside. $\endgroup$ Mar 7, 2017 at 21:33
  • $\begingroup$ I will try one more time: The production function is a mathematical object, not constrained by behavioral assumptions (like optimizing behavior). As a mathematical object, it may have regions where, if we introduce optimizing behavior, they will never be reached. But this does not make these regions non-existent. $\endgroup$ Mar 7, 2017 at 21:35
  • $\begingroup$ @Alecos I don't agree. To me a production function is a result of optimizing behaviour : what is the maximum output I can get with these inputs ? More formally, there are usually an infinity of feasible vectors in the set of production possibilities given the inputs, and the value of the production function for these inputs is only one of them. $\endgroup$
    – user11629
    Mar 8, 2017 at 18:57

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.