As automobile manufacturers are painfully aware, certain models experience slumps in sales. Inventory piles up, forcing manufacturers to find ways to stimulate sales. What are the options? In the United States auto manufacturers often turn to rebates, cash subsidies paid for each transaction. The two most commonly used subsidies are payments made directly to customers—“customer cash” in industry lingo—and payments made directly to car dealers—called “dealer cash.” Which one works better? New research by Meghan Busse (Associate Professor of Management and Strategy at the Kellogg School of Management) and Florian Zettelmeyer (Professor of Marketing at the Kellogg School of Management) shows that the answer is not quite as simple as it may seem.

Putting money directly in the customer’s hands may appear to be a sure way to lower the price the customer must pay the dealer. After all, every dollar the manufacturer gives the customer is a dollar the customer does not have to take from her own pocket to pay the dealer.

But there is a catch. “Prices for new cars are negotiated between car dealers and new car buyers,” Busse explained. “If a dealer knows that the customer is getting $1,000 from the manufacturer, the dealer may be able to get the customer to agree to a price that is $100 or $200 more than the customer would otherwise have agreed to pay.” As a result, the customer would still pay $800 or $900 less than he or she otherwise would have, but part of the rebate would go to the car dealer.

This might seem to weaken the allure of customer cash, but, as Busse points out, the effect cuts both ways. “If the manufacturer instead gives $1,000 to the dealer for every transaction completed, the dealer could pass along several hundred dollars to the customer in the form of a lower price in order to persuade the buyer to close the deal,” she said.

Customer versus Dealer Rebates
This research, published with colleague Jorge Silva-Risso (Associate Professor of Marketing at the University of California, Riverside) in The American Economic Review, finds that both car dealers and customers gain from manufacturer rebates, whether the manufacturers target the rebate to customers or to dealers. But determining who benefits most depends on which form of rebate the manufacturer chooses.

The researchers analyzed sales of more than 133,000 new cars sold in California from September 1998 through December 2000. Roughly 15 to 20 percent of the state’s new car dealers were represented. They studied the type of vehicle, the negotiated retail price paid by the customer to the dealer, the wholesale price paid by the dealer to the manufacturer, the value of any trade-in vehicles, and the dates and types of sales promotions. They also incorporated demographic information about buyers, such as education and income levels, and dealers, such as geographic region and the amount of local competition.

As is often the case when studying “real world” phenomena, the researchers had to clear some methodological hurdles. In particular, they needed to disentangle the roles of price and promotions.

Separating Price from Promotion
“The promotions may not be independent of price. There may be an overall price trend that leads manufacturers to offer rebates,” explained Busse. For example, an overall weakening of certain economy-wide features may cause sales to plummet, leading manufacturers to pull out all the stops to revive business. “To distinguish the impact of the promotion from the overall trend, we used a method called regression discontinuity, which is used a good bit in labor economics, but not so much in work like this.

“Basically,” she continued, “we looked at tiny slices of time, one week immediately before and after promotions began, changed, or ended. We compared the average prices during those two periods, which are so close together in time that the only major, discontinuous difference is in the promotion.”

The average negotiated retail price paid for a car was $25,490. Customers received cash from the manufacturer, an average of $1,242, in 26 percent of the sales. Manufacturers gave cash to dealers, $932 on average, in 18 percent of the sales.

When manufacturers offered cash directly to customers, the prices customers paid for their cars were lowered by roughly 70 to 90 percent of the rebate amounts. For example, a person given $1,000 cash by Ford when buying a new Taurus ultimately paid roughly $700 to $900 less than a person who did not receive any cash when buying the same kind of Taurus in a similar market.

But when manufacturers offered cash to dealers, customers only received roughly 30 to 40 percent of the benefit. That same $1,000 from Ford, when offered instead to the dealer who sold that same type of Taurus, would only lower the price the customer ultimately paid by $300 to $400.

Push versus Pull in Promotions
The practical implications of this work for the car industry are clear. “From a managerial point of view, what’s the most effective promotion, push or pull?” asked Zettelmeyer. “Give a supermarket shopper a coupon to encourage them to buy, to ‘pull’ the product off the shelves, or give a supermarket owner a discount to encourage them to sell, to ‘push’ the product off the shelves? For car manufacturers, both approaches stimulate sales in the wholesale market. But while customer cash is more effective at lowering prices, dealer cash may be used to help influence sales for a particular dealer who’s not doing so well.”

As the researchers point out, however, the work has implications that extend beyond the car industry. One of the key distinguishing features of buying a new car is price negotiation. In negotiations, economists (as well as many buyers and sellers who have had to negotiate a purchase) believe that little things can make a big difference to the outcome. One of those “little things” is having an information advantage.

“Customer cash and dealer cash set up a very interesting comparison,” Busse says, “because they are so similar in almost every way.” The key difference is that customer cash is heavily advertised, so that both the dealer and customer know that the manufacturer rebate is part of the deal. But only the savviest customers will have investigated whether dealer cash is available. “As a result, dealers have an informational advantage when manufacturers offer dealer cash rebates—only they know just how much money is on the table—while in negotiations involving customer cash rebates, both parties are equally well informed,” Busse concludes.

The effect of such “asymmetric information,” as economists call it, on the outcome of negotiations is something that has often been theorized and tested in experimental settings. But there has been little testing in real-world settings with large stakes. The research suggests that there may be many situations—buying a house, negotiating a salary, making a business deal—in which a slight advantage in information can lead to a big advantage in the outcome.