[Currents header graphic]

September 15, 1997

Economist says it's not quite as simple as Econ 101

Dan Friedman

By Jennifer McNulty

Economist Dan Friedman is the first to admit that his latest research interest has him heading upstream, but he has a sturdy paddle. A three-year $185,000 grant from the National Science Foundation assured him that the time is right for his provocative analysis of the performance of customer markets.

Customer markets, which include the labor market, retail, and wholesale markets, are characterized by transactions in which the parties care about each other's identity--workers care who their employer is and about conditions beyond salary, just as retail customers develop brand loyalties and other nonprice preferences regarding where they shop and from whom they buy.

Yet most of economic theory is based on auction markets, which, by contrast, are driven solely by price. The sale of stocks, bonds, and commodities like gold do not hinge on the identity of buyers and sellers. "With auction markets, the seller doesn't care who he's selling to, and the buyer doesn't care who he's buying from," said Friedman. "All that matters is price."

Friedman points out that only about a third of all transactions in this country are in auction markets; about two-thirds are in customer markets.

Friedman believes economic theory needs to be developed to recognize the complexities of customer markets. At the crux of his own work is Friedman's desire to predict when people will change transaction partners--a complicated question when one considers the costs of investigating a job change, for example.

Friedman is convinced that economists embrace the auction market model more broadly than they should because it generates sharp and striking conclusions--whether or not it is a good fit.

For example, Friedman said that if the labor market worked the way it's "supposed to," unemployment wouldn't ever be a problem. In an auction market, a drop in the demand for labor would immediately lead to a decline in wages, and job loss would be minimal. "In reality," he said, "wages remain steady and workers are laid off, creating unemployment."

Unemployment is just one illustration of phenomena that are not explained well by current economic theory. Such phenomena are described as "frictions," and Friedman said it's time to take them seriously.

"Auction markets are efficient and frictionless, but they don't fit human nature," said Friedman. "They don't encompass human features such as fairness and community, and those considerations come into play in many of our transactions. Existing theory is a useful starting point, but we've got to take the next step."

Friedman is developing a theory that accounts for the frictions that are commonplace in customer markets. "Think of these frictions in terms of what economists call switch costs--for example, how much extra pay it would take for an employer to induce you to make a job switch," explained Friedman. "You wouldn't take a job in New York City just because it paid 10 cents more an hour, right? Even after taking into account the different housing prices, it costs a lot to move. You have to ship the furniture and, even more costly, you have to adjust to your new surroundings. To make it worth your while, the employer would have to pay you a lot more than you make here."

Understanding how switch costs impact markets has implications for economic policies, particularly regarding labor, which typically has the highest switch costs. "Government could do things to reduce or increase switch costs," said Friedman.

Some believe the reason the U.S. has fared better than Europe in the 1990s is because we have lower labor-market frictions and our unemployment benefits do not require workers to stay in the same industry or geographic area. Social democratic governments, such as Sweden's, may embrace higher switch costs--lowering incentives to relocate, for example--to help preserve family stability, he said.

More sophisticated knowledge of customer markets also has implications for monitoring price dispersion. When identical goods are selling simultaneously at different prices in different locations, the traditional interpretation is anticompetitive or monopolistic behavior, noted Friedman, but the explanation in some cases is simply switch costs.

In developing a theory of customer-market performance, Friedman is using sophisticated new analytical techniques that have become available to economists only in the last couple of years. His goal is to develop a theory that predicts when it is worthwhile for customers to make a switch--whether it's changing jobs or laundry detergent.

"The customer's problem is a bit like an investor's problem in deciding when to exercise an option to buy shares of stock," said Friedman. "I think I can use the same mathematics--a certain partial differential equation--to help predict how customers respond to price changes and how they respond to changes in search costs or the availability of other trading partners."

To test his theory, Friedman and his team of researchers will conduct lab experiments in which individuals interact as buyers and sellers. In the experiments, researchers will manipulate "switch costs" from low to high to mimic the circumstances of auction and customer markets. To elicit genuine behavior, Friedman will have the subjects exchange real money and keep any profits they earn during the transactions.

"We should be able to look at the outcomes and see which are due to different switch costs," he said. "It will be a much more direct measure than you can get out in the world. The theory isn't all in place, but there is progress that can be made."

Friedman is eager to see how well his theory performs in the lab. And if he finds that it only explains 50 percent of the switches customers make, well, he's willing to study the impact of noneconomic factors that affect customer decision making, such as fairness, loyalty, and emotional disposition.

To the Currents home page

To UCSC's home page