Paul F. has been selling bagels and doughnuts at nonprofit research firms in the Washington, D.C., area for more than 20 years. Paul was an economist before that, and a good one. (I have known him since 1963.) When Paul reached his early 50s he was able to retire from the research business and devote himself to a less demanding vocation as The Bagel Man. Joshua Gans writes about Paul in a post at Aplia Econ Blog entitled "Maximizing the Bagel Dollar." There, Gans summarizes a paper by Steve Levitt (abstract here, paper available by subscription only), in which Levitt analyzes the wealth of data collected by Paul in his years as a bagel and doughnut vendor.
Paul buys his wares fresh daily and then schlepps to various locations around the D.C. area, where he offers a variety of bagels (with cream cheese on the side) and doughnuts. He delivers his wares to each site once a week -- an event known as Bagel Day. Paul doesn't hang around to collect money from his customers; he trusts them to leave the right amount of money in a collection box, and his trust is generally well-placed.
After making his deliveries, Paul cycles back through his route to pick up his collection boxes. He then meticulously records his sales, overages, shortages, left-overs, and stock-outs. Levitt's analysis is based on Paul's meticulous records. According to Gans,
Levitt shows that Paul gets quantities right (for both bagels and donuts, there are very few stock-outs or left-overs), but he often gets prices wrong. Paul systematically prices donuts and bagels too low. Indeed . . . Paul could actually increase total revenue by increasing his prices. Hence, his prices are at a point like P in the figure. At point P, marginal revenue is negative (below the x-axis); so increasing price and reducing sales will raise revenue.
The question is whether Paul -- the savvy economist -- really sets his prices "too low." I suspect not. Paul doesn't need the money. Knowing Paul, he's making himself happy by more-or-less correctly judging the number of bagels and doughnuts he can sell at prices that don't cause him to lose money. That is, he wants to maximize the quantity of bagels and doughnuts enjoyed by his customers, as long as he's making "enough" to warrant the effort he puts into his business. There's more to income than money, and Paul's behavior is a good case in point.
The kind of behavior exhibited by Paul also is a good long-run business strategy. "Underpricing" helps to build loyalty. Consumers are less likely to switch to competing vendors and products if they have been pleased by a vendor's or manufacturer's long record of offering good value for the money.