Sunday, July 15, 2012

Hal R. Varian: How to Build an Economic Model in Your Spare Time


I googled economic model and Hal Varian's PDF (16 pages ) showed up, second on the list.

The first step is to get an idea. This is not all that hard to do. The tricky part is to get a good idea.

I have a good one.

So let's assume (a favorite word of economists) that you have an idea. How do you know if it is any good? The first test is to try to phrase your idea in a way that a non-economist can understand. If you can't do this it's probably not a very good idea. If you can phrase it in a way that a noneconomist can understand, it still may be a lousy idea, but at least there's hope.

Pretty good paper. Funny in spots.

I think I'm pretty good at putting it simply.

Before you start trying to decide whether your idea is correct, you should stop to ask whether it is interesting. If it isn't interesting, no one will care whether it is correct or not.

I'm sure this is correct, and pretty important. I have trouble with it. I thought, if you build it they will come. I thought build a better mousetrap, they'll beat a path to your door. (I didn't really think that but... you know.)

So let's skip the literature part for now and try to get to the modeling. Lucky for you, all economics models look pretty much the same. There are some economic agents. They make choices in order to advance their objectives. The choices have to satisfy various constraints so there's something that adjusts to make all these choices consistent. This basic structure suggests a plan of attack: Who are the people making the choices? What are the constraints they face? How do they interact? What adjusts if the choices aren't mutually consistent?

Asking questions like this can help you to identify the pieces of a model. Once you've got a pretty good idea of what the pieces look like, you can move on to the next stage.

See, here's where I get in trouble. I think Hal Varian is talking about micro models. So I want to ignore this part of his advice. Micro models are never going to solve the macro problem.

I think what Hal describes as a model is really just an explanation. According to him, a model tells who makes the choices, what the constraints are, how they (not sure who) interact, and what adjustments are necessary. Sounds like yadda yadda yakyak to me. Micro yakyak.

I don't think Keynes made Hal Varian-style models in The General Theory. I don't think Adam Smith made Hal Varian-style models in The Wealth of Nations. I think I do things like Smith and Keynes. Ego, I know. So, tell me what you think. But be specific. I don't understand hints.

A model is supposed to reveal the essence of what is going on: your model should be reduced to just those pieces that are required to make it work.

Okay.


On page 10, Varian's theme degenerates into "Planning your paper". It's not a model anymore now, it's just a paper. It goes on for another half a dozen pages and never goes back to being a model.

So, okay. A "model" is not necessarily a computer simulation of the economy or anything involved and complex like that. A model is an explanation.

// Part 2 of 4

3 comments:

Jerry said...

I think that (sometimes?) "a model" might mean: a simplified version of the truth, that sort of gets at the important parts while ignoring everything else.

For instance, a map of new york doesn't contain all of the details about where each sidewalk, or hot-dog vendor, or bump in the pavement is. If you wanted to include all of the detail, the map would be as big as the city. And not as useful.

Or, the C-language has a "model" of a computer that it works with. (I mean basically the way memory is handled). It's not a perfect fit - e.g. there isn't anything that talks about on-chip caches.

But, it's a good model because it's close enough to true to give you good control over what's going on, but you (usually) don't have to write different code for different chip architectures.

So, I don't know what that means. Maybe an explanation, yeah. A statement about what you think the important / driving pieces of it are. Of course, there are more pieces than that. But if you picked the 1 or 2 or 3 most important pieces and ignore the rest, the system might become simple enough to use to draw some useful conclusions or guidance about what to do.

Tom Hickey said...

In philosophy of science a theory is an explanation. A scientific theory is formalized so that theorem can be deducted from the starting points (assumptions in a theory about how things stand) according to formation and transformation rules given by logic and math.

Models are logical constructs for mapping the explanation onto how things are claimed to stand in the world. The assumptions and theorems generate hypotheses that can be tested against observation of how things stand. This permits the semantic interpretation of the theory as a scientific one. Otherwise the structure is remains formal and analytic. Math models stays nothing about how things stand unless they are interpreted semantically in terms of a representative model. This requires mapping.

So there is explanation — model — observable world of events. The observable world provides the data that become information for the model. The design of the model acts as the mesh of a net for capturing data and structuring information in a useable way for the approaching the design problem through mapping the explanation and checking it against that which is being mapped based on feedback from observation.

HMN said...

Re:" Observable world events"

This is where we some people run into dynamic problems! Perhaps your model will need some variables that represent valuations ... LOL!

==> The FASB describes Level 3 inputs as “unobservable.” If inputs from levels 1 and 2 are not available, FASB acknowledges that fair value measures of many assets and liabilities are less precise. Within this level, fair value is also estimated using a valuation technique