Sunday, December 21, 2008

Rationalize all you want

Think of the most rational person you know. If a friend comes to you with a problem, would you open a spreadsheet document? Would you type Pros at the top of Column A and Cons in Column B? Then I bet you would ask your friend to list all the pros and cons. Maybe you would even ask your friend to assign point values to each item listed. Then you sum up each column. The side with the higher point value would be the answer to the dilemma!

We would describe this kind of rationality as logic or objectivity. Economists assume that, on average and in general, people make decisions rationally. There are exceptions, of course; people who seem to be always irrational or moments in an otherwise rational life when crazy things happen. We know that people aren't always rational. Or at least they don't appear to be rational.

But let me clear what I mean by rational. Remember that economics is all about making choices, such as whether to buy something or not. Rationality means that you would figure out the value of the benefit from the purchase and the value of the cost. Then you would compare the benefit to the cost. If the benefit is greater than the cost, then you would make the buy. If the benefit is less than the cost, then you would not purchase.

Notice how vague the terms benefit and cost are. Benefits and costs will vary from person to person. Benefits and costs may even vary within one person over different periods of time. Emotions can fit under this definition of rational. If our measurement of the benefits includes emotion, we can still act rationally given that we compare that emotional benefit to however we measure the cost. You might call this subjective rationality.

When we see someone acting supposedly irrationally, it may be that we just don't understand how that person is measuring benefits and costs.

Rationality also means that, on average and in general, people make decisions consistently and predictably. Here's an example. How many readers out there like apples better than bananas? Raise your hands and keep them up. There are 20 of you with hands up. Now, how many of you 20 like bananas better than cantaloupes? Keep your hands up. Cantaloupe lovers can put their hands down. OK, there are seven hands left up.

Now, for the rest of you with your hands down, can you make a prediction about these seven people with their hands up? What would you predict these people would answer if I asked if they liked apples better than cantaloupes? You would predict that all seven would like apples better. Why?

You would predict that they would prefer apples over cantaloupes because of the principle of transitivity. If someone likes apples better than bananas and bananas better than cantaloupes, then that person should like apples better than cantaloupes. If A > B and B > C, then A > C. We expect people to act in this consistent and predictable way rather than to answer randomly about fruit preferences. This is another part of rational behavior. When making choices, people do not behave randomly, but instead they make decisions, on average and in general, in predictable ways.

This assumption of rationality is comes from the fact that economics is a social science. As such, economics uses the scientific method. Economists observe human (and sometimes animal) behavior and then ask why does this behavior occur. But different from the physical sciences, it is difficult to conduct economic experiments. Possible, but difficult, and subject to limitations. Almost all economic behavior occurs in real life. It's hard to ask customers in a store to divide into a control group and an experimental group. Therefore, most economic experiments are thought experiments, often reliant on mathematical theory and statistical analysis.

Nevertheless, economists do construct scientific models. There are four parts to a model:

1. assumptions
2. isolation
3. story
4. prediction

Every model begins with some form of the verb to assume. The purposes of the assumptions are to simplify and control the circumstances of the thought experiment. Real life is complex and messy. So, one assumption might be to freeze frame everything, hold everything constant. This eliminates the effects of factors that are beyond the scope of the experiment. For example, we might assume that incomes stay constant while we are studying shopping behavior.

Assuming that everything is held constant, we can then isolate on the factors to be studied. And in many models, especially at introductory levels, we isolate on two things. Why two? Because in economics, everything comes in two's! For example, we might isolate on price (independent variable) and quantity demanded (dependent variable).

A story is a sequence of events, a chain reaction. One thing leads to another. We have isolated on two things in our model and the story begins, "Change one of the things." For example, we might change the price by increasing or decreasing it. Changing the price sets off a sequence of events.

And that sequence of events leads to a conclusion, an ending to the story. Making the same assumptions, isolating on the same two things, and telling the same story, we expect that the ending will always be the same. Therefore, the conclusion becomes a prediction: If these things are true, then we predict that this result will occur.

Economic models help us predict human behavior. A scientific model that accurately predicts real-world behavior is called a theory. Economists never say, "Oh, that sounds good in theory, but it would never work in practice." If it doesn't work in practice—if it doesn't predict real-world behavior—then it isn't a theory. The idea is instead a failed hypothesis.

The basic economic models presented in this blog have been tested by time and do a good job of predicting real-world behavior.

Note to behavioral economists
Advocates of behavioral economics question the assumption of rationality. They criticize many economic models and equations that primarily revolve around benefits and costs measured in money. Behaviorists believe that most economic models assume that all people consistently behave the same way all the time. Predictable behavior based on money might be called objective rationality.

Objective rationality, however, applies only to simple economic models with one or two independent variables. The same models can be expanded by adding parameters and still meet the standards of marginal analysis and transitivity. This is the foundation of subjective rationality.

COPYRIGHT © 2008 by Robert D. Sandman
ALL RIGHTS RESERVED.

No comments: