Psychology and Behavior

When Business Competition Harms Society

In highly competitive markets, many firms are likely to bend the rules if doing so will keep their customers from leaving for a rival, according to new research by professor Michael W. Toffel and colleagues. Case in point: service stations that cheat on auto emissions testing.

Last winter, a senior admissions officer at Claremont McKenna College resigned, after admitting to inflating reported SAT scores of the incoming class for six years and sending the falsified reports to U.S. News and World Report.

It's very cheap to get a second opinion.
—Michael W. Toffel

The scandal spawned widespread questions about the value of competitive rankings—and even about the value of competition in general. Was the officer's behavior understandable? Did intense competition naturally lead to unethical behavior?

A new research paper, Customer-Driven Misconduct: How Competition Corrupts Business Practices, suggests that many organizations in highly competitive markets are likely to bend the rules if doing so will keep their customers from leaving for a rival firm.

"Competition is generally thought to be good for economies because it keeps prices low and quality high. But when meeting customer demand is bad for society at large, then competition has a flip side," says Victor Bennett, an assistant professor at the USC Marshall School of Business, who cowrote the paper with Lamar Pierce of Washington University Olin Business School, Jason A. Snyder of the UCLA Anderson School, and Michael W. Toffel of Harvard Business School.

In the quest to discover whether competition breeds unethical behavior, the researchers examined the vehicle emissions testing program in New York State, one of several states required by the federal Environmental Protection Agency to institute such a program. The researchers examined the prevailing inspection system called On-Board Diagnostics II (OBD-II), which tests for excessive carbon monoxide, hydrocarbons, and nitrogen oxides, and also checks the power-train systems for problems that might eventually lead to elevated emissions levels. Vehicles that fail the annual test are banned from being driven in New York until they undergo necessary repairs, or are sold to a buyer in a state that does not require emissions testing. (New Yorkers do receive a one-year waiver for cars requiring repairs that exceed $450.)

Rather than operating its own testing facilities, New York outsources its emissions testing program to the private sector—to service stations, garages, and dealerships—but mandates they charge customers the same fixed price for the service, a nominal $27 for tests within the New York City metropolitan area and $11 elsewhere in the state. Because these facilities don't have the option of competing for business on the basis of price, they compete solely on the basis of quality.

"With vehicle emissions testing, high-quality service from a customer's perspective often means 'the inspector passed my vehicle' rather than 'the inspector accurately measured my vehicle's emissions,'" says Toffel, an associate professor and Marvin Bower Fellow in the Technology and Operations Management Unit at HBS. "This results in facilities competing on the basis of leniency to attract and retain customers. Clearly, the government has quite a different perspective as to what constitutes a high-quality emissions test."

Importantly, cars that fail an emissions test at one facility are allowed to be retested at another facility. "It's very cheap to get a second opinion," Toffel says. "It's certainly cheaper than paying for repairs."

The study shows that owners of vehicles that fail an emissions test are 11 percent less likely to return to that facility the following year, which implies that those same customers also are less likely to use that business for future repairs and routine service.

"The short-term benefit of failing a vehicle pales in comparison to the long-term benefit of retaining the customer's service and repair business," the paper explains. "Annual service and repair expenses on a 2006 Jeep Grand Cherokee, for example, are estimated at over $2,200/year by Edmunds.com, with gross margins on such repairs typically at 50 percent."

Hence, testing facilities have ample incentive to exhibit leniency, meaning, they manipulate the testing equipment or find other ways to pass vehicles that should fail to keep their customers loyal. So it's no surprise that the emissions testing industry is fraught with fraud all over the country.

"California, New York, Georgia, Nevada, and many other states have had scandals involving lenient vehicles emissions testing in the past few years," Bennett says. "It's an ongoing, enduring problem."

Testing The Testing Facilities

To learn whether increased competition was associated with a greater likelihood of testing results being fudged, the researchers looked at pass/fail results in micro-markets encompassing a 0.2-mile radius. They compared markets with very few testing operations against those micro-markets that were peppered with facilities, reasoning that the more facilities in any given area, the higher the competition—and the higher the incentive to bestow leniency on customers.

"We considered how easy is it for customers to leave one facility and go to another one if they're upset that they're not being passed," Bennett explains. "New York is a good place for this type of research in that you've got a lot of diversity, from rural areas where facilities are miles and miles apart, to places in the Bronx where you'll see eight or nine of these facilities on a single block."

The New York State Department of Environmental Conservation provided the research team with the records of all 28 million OBD-II inspections conducted on gasoline-powered consumer vehicles at the state's 11,425 license testing facilities from 2007 to 2010. Researchers compared the pass/fail records of similar vehicles in both densely populated areas and sparsely populated areas, paying close attention to older vehicles with high mileage—that is, those more likely to need repairs in order to meet emissions standards.

The data analysis showed that vehicles were much more likely to pass if they were tested at a facility that was located near a competitor. Thus, a jalopy had a better chance of passing at a garage at an intersection or on a block that had many testing stations compared with a lone garage located a half-mile away. Competition for business seemed to be the clear reason why.

"Consider two vehicles of the same make, model, and usage, both tested in the same month in the same city. You'd hope the only thing that would sway whether these vehicles passed would be how well they were taken care of," Toffel says. "But what we find is that the one tested at a facility with more nearby competitors is more likely to pass. An extra facility within two blocks of another seems to lead to the passing of one out of every 100 cars that shouldn't pass."

Pass rates were found to be highly inflated at facilities that had recently entered the vehicle inspection market, especially if they were located near older, more established facilities. "To be able to make it in a new market, you have to offer something really compelling," Bennett explains. "And in this market what you can offer to your customers—the thing that will draw them away from the competition—is passing their car."

Managers Take Note

The upshot is that firm misconduct appears to increase with the threat of losing customers to rivals, which has implications for both business managers and government regulators.

For government agencies, the findings indicate the importance of monitoring firm behavior in highly competitive markets—and the importance of realizing when a policy is what creates intense market competition in the first place.

"If regulators care about the efficacy of these policies, they should be oversampling their inspections in highly competitive markets because that's where the problems are most likely to occur," Toffel says.

(In related research, Pierce and Toffel have investigated two other factors associated with leniency of vehicle testing facilities: the other business lines these facilities are engaged in, and their governance and ownership structure. They found that facilities are more lenient when their organizational scope includes products and services, like selling cars, where enticing customer loyalty can enhance profits. On the other hand, compared with independent facilities, subsidiaries or branded affiliates exhibit much less leniency. See their working paper, The Role of Organizational Scope and Governance in Strengthening Private Monitoring .)

For Future Study

Next, the researchers plan to study whether government crackdowns on unethical testing shops encourage competitors to clean up their own behavior—or just the opposite.

"We can imagine one of two things happening," Bennett says. "One possibility is that if you see your competitor get caught in a high-profile sting, then you become afraid that you're next, and so you get scared straight. But the other possibility is that you realize that there are all these customers who used to go to the unethical facility, with the intention of being passed even though they didn't deserve it. You've got this big orphaned market of customers looking for that. So all of a sudden you might become more lenient to try to attract them—until you get caught. For want of a better analogy, it could be like a game of Whac-a-Mole."

Latest from HBS faculty experts

Expertly curated insights, precisely tailored to address the challenges you are tackling today.

Strategy and Innovation

Social Responsibility

Diversity and Inclusion