Tuesday, February 26, 2013

Excellent sentences from Don Boudreaux

"Mr. Forester is exactly correct that ceteris paribus seldom, if ever, holds true in reality – especially in social and economic reality. And this fact is precisely why ceteris paribus theorizing is so crucial. Our theories are meant to give us mental pictures of parts of reality. To take mental snap shots of any part of reality that we wish to better understand, we cannot rely upon reality itself to isolate for us those parts of itself – those parts of reality – that we’re most interested in understanding. So we theorize, and in theorizing we economists (as do all scientists) use ceteris paribus restrictions. To use such restrictions is no sign that the scientist really believes that ceteris is paribus - or that ceteris should be paribus. To use such restrictions is simply an unavoidable part of thinking seriously about an ever-changing and complex reality.

Mr. Forester is also correct that (and here I use economists’, not his, terms) partial-equilibrium analysis differs from general-equilibrium analysis. Understanding a part of the economy, working in isolation, is not necessarily to understand that part of the economy as it is connected to the larger economy. What is true for that part of the economy M were it isolated from the rest of the economy might well not be true for that part of the economy M when the effects of changes in M extend outward to other parts and then, in turn, the reactions from the other parts of the economy work their way back to affect M.

But in order to legitimately dismiss the lessons of partial-equilibrium analysis because of alleged contrary feedback from the larger, general economy, one needs more than just-so stories or hypotheticals."
The first two paragraphs are more highly recommended than the third here. I have no idea why he thinks we ought to privilege partial over general equilibrium theories. I also don't see the general equilibrium stories that get told as being any more "just so stories" than the partial equilibrium stories out there. In fact if anything people are too quick to stop thinking at the end of a naive partial equilibrium story.

I share Don's skepticism, later in the post, of the idea that general equilibrium is what is driving observed positive or non-existent minimum wage effects on employment. I said this the other day, too. A general equilibrium story that says minimum wages help the whole economy is more plausible than that. But the strongest theoretical explanations for the emploument effect are partial equilibrium ones in this case I think.

We have a few good partial equilibrium stories - some that predict positive effects and some that predict negative effects. Good empirical design handles the ceteris paribus, and what comes out helps us to arbitrate between our theoretical possibilities.


  1. Daniel,

    Raising the minimum wage is no more an economic issue than ending slavery.

    Doing either increases the cost of business and will likely result in fewer people being employed.

    To be truly effective, our minimum wage laws should also apply to products or services produced elsewhere, but offered for sale here. Again, what is the difference between a certificate that a product was not made with slave labor and a certificate that the product was made by someone who was paid a living wage

  2. I'm not an economist, but it seems as if taking on gargantuan tasks such as building a general equilibrium model from microfoundations would seem like it needs to be done in parts.

    I.e. Take the basic supply and demand model, and note that an increase in income shifts the demand curve towards the right and a decrease in prices shifts the supply curve to the right.

    Then link together many different supply and demand models together under the assumption that each individual agent has a limited income and needs to purchase a limited number of products (each with their own markets) and produce other products in order to survive.

    So the way I see it, partial equilibrium models are like jigsaws, and the underlying assumptions are the links to other jigsaws.


All anonymous comments will be deleted. Consistent pseudonyms are fine.