Saturday, July 07, 2007

Decreasing Marginal Costs Empirically

In my last post I linked to Debunking Economics: The Naked Emperor of the Social Sciences. When I first encountered the book I hadn't taken a class on economics. I glanced through the book, noticed the models, and tossed it aside - I don't like studying models unless I have to. But now I know that I will have to check it out next time I'm at my school's library.

Anyway, the AmazonConnect author's blog is very interesting. His latest post claims that the empirical reality of the firm doesn't match up with the theory. He reports that firms face decreasing marginal costs rather than the standard "always increasing" marginal costs. That hugely invalidates theory, yet, as I commented it makes no sense. A firm producing with fixed capital assets can only produce so much before they are constrained by their limited fixed capital - thus driving the price up. When the fixed costs are increased, the marginal costs fall to a workable level again. Instead I think he may be confusing average and marginal cost. As firms increase production without increasing fixed costs they will spread the fixed costs out. Most people have heard of this: it is called economies of scale and it is a huge force at work in industries today ranging from semiconductors, cars, pharmaceuticals and even software production (wherein the fixed cost is the initial time/money spent on the program and MC is the cost of replicating the software for sale -- essentially just packaging costs). Economics of scale will sometimes produce a natural monopoly.

It wouldn't surprise me if most firms faced fairly constant marginal costs up to a threshold level (that level being the capacity of a plant or the capacity of a firm to provide services without hiring more employees). At a certain point one's employees and one's factory becomes overextended. One has to work people overtime and run the machines overtime. Ultimately marginal cost must increase if fixed costs remain fixed. Correct me if I'm wrong.

However, empirically most businesses will fluidly increase their "fixed" costs as MC begins to increase, so businesses wouldn't notice increasing marginal costs. But they certainly notice decreasing average costs.


Anonymous said...

Have you thought of a video store that can have decreasing MC as each movie is rented out more than once, Free to air television has decreasing MC until they reach full capacity and need to increase coverage. Housing rental the fixed cost is the same but the MC of renting for another day can be zero if there is no damage to the house.

Tomas Hagberg said...

The marginal cost that a producer of a refined good meets is deminishing as long as it's demand does not increase the price of the inputs. This is because the economies of scale exceeds the crowding costs. In reality the company has lower marginal costs of producing one extra product than producing the last one... I am aware that I contradict the oldfasioned autistic economical theory, but it has to be done in order to understand reality...

undergroundman said...

I haven't had much to eat, so I'm not feeling very sharp -- and I'm not following at all what you're saying. As long as demand does not increase the price of inputs?

I think economic theory accepts economies of scale, which is what it sounds like you're saying...but economics points out, correctly, that increasing production holding the size of the factory equal, you experience increasing costs. If you expand the factory without expanding other things, you also get these increases. Now, if you expand everything, you can get economies of scale -- but eventually inputs become scarce (unless the elasticity of supply is very high -- but even then, eventually primary inputs will become scarce).

Tomas Hagberg said...

The point is that it is unrealistic to hold the size of the factory constant in any way regarding to the prize mecanism. The number of potential factories is virtually infinite. Maybe in a few hundreds years you will be right. But at least not for now.

Though, it would be realistic holding for example the size of an agricultural area constant. This depends on the crowdedness leading to scarcity.