Blog Post

Examining the BP Deepwater Horizon Oil Spill and the Making of a Consumer Boycott

Feb 26, 2015 | Zhongmin Wang

Information on consumer boycotts and their participants is highly relevant for many corporations. A 2005 survey found that 36 percent of consumers polled across 17 countries had personally boycotted at least one brand, a likely troublesome figure for marketers. But what makes a boycott worth noting for companies or observers? To help answer that question, my colleagues, Alvin Lee and Michael Polonsky of Australia's Deakin University, and I created a model for market-level boycott intensity that distinguishes between the levels of egregiousness—an event’s severity—and the preferences consumers associate with them, something that the existing literature on consumer boycotts fails to address. In a new RFF discussion paper, “Egregiousness and Boycott intensity: Evidence from the BP Deepwater Horizon Spill,” we also explore the relationship between boycott intensity and media coverage, due to the latter’s ability to influence consumer awareness about egregious events.

In order to test our model empirically, we turned to data collected during the Deepwater Horizon oil spill. The spill represents a single significant event—but its egregiousness changed over time based on reported levels of leaking oil. It also provided us with an easy way to conceptualize the distinction between egregiousness levels and consumer perceptions—the oil spill’s overall severity didn’t change after it was stopped, but consumers’ negative feelings of the event declined over time. Using our market-level model, we determined an area’s boycott intensity by measuring decreases in BP brand’s market share across 257 US counties.

We expected that the oil spill, at any given level of egregiousness, would have caused greater harm in markets closer to the Gulf Coast region and in markets with stronger environmental preferences—a hypothesis that was confirmed by our model. Our findings also indicated that while drivers boycotted BP gasoline as the oil spill was occurring and that boycott intensity increased with the official flow rate, this pattern disappeared quickly after the spill was stopped. In terms of the media’s influence on consumers, boycott intensity appeared to increase with the weekly number of spill-related news hits, suggesting that consumers may have made assumptions about the seriousness of the spill due to the intensity of the media coverage it received at any given time.

Ultimately, our results suggest that variations in consumer characteristics can magnify into market-level variations in boycott intensity. This relationship may prove useful to companies interested in mitigating the impacts of a future firm-specific boycott, as they may be able to benefit by adopting market-based strategies to counter these effects. However, the conclusions we draw about boycott thresholds and intensities suggest that a firm’s best defense against large-scale consumer protests is to adopt measures to avoid highly egregious events in the first place.