Demand shocks are one of the most basic features of economic theory, representing a sudden rise or drop in consumers' desire to purchase a good or service. But empirical evidence of how companies and industries respond to demand shocks is harder to come by. Thomas Hubbard, a senior associate dean of strategic initiatives at the Kellogg School of Management, found inspiration for a model of demand shocks in his own memories of childhood family vacations.
"When I was a kid riding in the car to Florida, I-95 wasn't completed yet, and we had to take side roads," he says. "That made me wonder recently what happened to the gas stations in towns when the interstate highway came through. I realized that these were demand shocks, and that I could answer a more general question by studying them."
Hubbard's paper "The Economics of 'Radiator Springs': Industry Dynamics, Sunk Costs, and Spatial Demand Shifts," coauthored with Jeffrey Campbell of the Federal Reserve Bank of Chicago, takes its name from the Disney animated film "Cars." In that film, local businesses in the fictional town of Radiator Springs suffer a demand shock when a new interstate highway bypasses the town.
“9/11 was a demand shock for the security scanner industry; you also see these shocks in the wake of a particularly effective advertising campaign,” Hubbard explains. “But these are single data points: they only happen once, and they happen everywhere at the same time.”
But in the real world, the effects of demand shocks can be difficult to study because of the singular events that often precipitate them. "9/11 was a demand shock for the security scanner industry; you also see these shocks in the wake of a particularly effective advertising campaign," Hubbard explains. "But these are single data points: they only happen once, and they happen everywhere at the same time."
A Textbook Example
Studying demand shocks to gas stations during the construction of the interstate highway system proved advantageous because "they're observable many times over many years in many regions," says Hubbard. "The government systematically collected data about the size and number of gas stations in every county every year since the early 1960s. I combined that information with another dataset showing when every single mile of the interstate highway system opened."
Because he wanted to compare the effects on gas station demand depending on whether the interstate highway opened up near or far from existing routes, Hubbard also had to "go back with maps to figure out what the best way to get between two cities was in the 1950s, then figure out where the new interstate ultimately got built, and measure the distance between the two." The results surprised Hubbard, who expected to find many examples of interstate construction occurring miles away from older highways (as in the case of Radiator Springs).
"What I didn't realize," he says, "was that especially in the western part of the United States, which wasn't densely populated, the interstate was more or less a widening of the existing road. That turned out to be very useful, since it gave me more variation than I expected in the spatial dimension of these demand shocks to gas stations."
Hubbard's findings empirically confirmed a "beautiful, textbook example of what economic theory would suggest": when interstate highways were built very close to existing roads, nearly all of the adjustment was in the size of firms, rather than the number of firms—gas stations responded to the increased demand by expanding in size and hiring more employees, but few additional stations were built. Meanwhile, when a new interstate route was opened several miles or more from the previous highway, entirely new gas stations opened up to service the demand, but the size of existing stations stayed the same.
The results suggest that entry opportunities in expanding markets aren't as simple to exploit as they might seem. "You have to pay attention to where the demand growth is happening," Hubbard says. "If you have more customers demanding the same stuff that they did before—like more drivers stopping for gas on an interstate very close to an existing highway—this doesn't represent a great chance for new firms to come in. If you want to meet that additional demand, you're probably going to have to replace one of the existing firms, not coexist with it."
At the same time, Hubbard's findings make a statement about sunk costs that the makers of "Cars" might appreciate. "Disney actually got it right," he says. "The film portrays a situation in which the new highway is built far from Radiator Springs, which makes the businesses in town suffer. But they aren't dead; they don't exit the market. Our data showed that after a similar demand shock, new gas stations opened, but existing ones didn't close—probably because they're sunk costs. You can't turn an old gas station into anything else, so you might as well keep it running, even if the new demand is happening elsewhere."
Related reading on Kellogg Insight