20 Years of Work Used to Fine-Tune Economic Model

Article excerpt

CAMBRIDGE, Mass. _ Remember the model planes your neighbor's brother built as a kid?

With a little glue, time and patience, he transformed dozens of tiny pieces of plastic into a work of art.

Talk about patience, Jay Forrester of the Massachusetts Institute of Technology has spent 20 years working on another kind of model: Gluing tens of thousands of pieces of information together with 2,200 mathematical equations that he feeds into a powerful computer.

Just as model planes simulate real ones, the 73-year-old scientist's model is designed to simulate the real U.S. economy.

He thinks it works better than models used by commercial economic forecasters, mainly because in building it, he relied more on engineering principles than economic theories.

Forrester, who developed one of the first general-purpose computers, set out in the early 1950s to discover why some businesses succeed and others fail.

After studying inventory control problems at General Electric Co., he deduced that the company's internal decision-making practices _ not outside influences _ governed its success.

He soon founded the field of "system dynamics," which is based on the notion that the structure of a company, or an economy, leads to success or failure.

In short, people respond predictably to circumstances that are dictated by the structures around them. For example, every company has a chain of command, and a middle manager's decisions often are influenced by what he or she thinks the boss wants. In the same way, the structure of the economy also leads to predictable behaviors.

So instead of programming his computer to look for specific correlations between things such as interest rates and car buying plans _ which is what economists do _ Forrester decided it would be better to study decisions and try to learn how decisions by thousands of different people ricochet off one another.

He devised equations linking likely decisions of managers in banks, factories, retail stores, government agencies and households.

Economic relationships, he says, are not automatic and often defy logic. Just because interest rates fall doesn't necessarily mean masses of people will rush out to buy cars. Such decisions also are based on other factors, such as job security, which is affected by corporate management policy, which is influenced by government policy. …