The automobile circa 1895 was perceived as a new, dangerous invention and a plaything of the rich. So the earliest auto races were essentially reliability trials, which morphed into mass-marketing opportunities for competing automakers. As Henry Ford recounted in 1922: “That ‘Model B’ — the first four‐cylinder car for general road use — had to be advertised. Winning a race or making a record was then the best kind of advertising.”
Auto racing’s first few decades created a historical laboratory for Hayagreeva Rao’s 1994 “Social Construction of Reputation: Certification Contests, Legitimation, and the Survival of Organizations in the American Automobile Industry: 1895–1911.” Rao is currently a professor of organizational behavior at Stanford University, and his paper was the first to depict market reputation as socially constructed.
Some 20-plus years later, management and entrepreneurship professor Brent Goldfarb at the University of Maryland’s Robert H. Smith School of Business has co-authored a “robust replication” of the study. It comes with a backdrop of management researchers suggesting anti-replication bias within their field is placing “management theory at risk of being more like a house of cards than a strong and enduring edifice of tightly welded steel beams.”
Goldfarb, with Anastasiya Zavyalova (Smith PhD graduate and Rice University professor) and Sandeep Pillai (Smith PhD candidate), reconstructed Rao’s data to reaffirm the main correlation reported in the original paper: U.S. automakers’ cumulative race — or certification contest — victories were associated with staying in business. They further confirm that startups experienced no advantage among relatively established peer firms. They expand on the original, noting that placing second, third or merely participating in races at all also predicts firm survival.
For example, one of the datasets shows 1,176 different automakers having a two-year median length of survival, with 72 percent of those firms failing during the study period. For the firms in this group that participated in races, those numbers are 4.5 years of survival and a 35-percent failure rate.
However, the study points out that “correlation does not imply causation and the empirical setup of both the original study and replication study lack a measure of reputation,” Goldfarb says. And likewise, “a series of presented tests appear more consistent with a mundane explanation: more capable firms were more likely to have their cars entered in races and also more likely to survive.”
The paper, he adds, “outlines the assumptions necessary to believe the idea that race victories led to a meaningful difference in firm survival.” The research also considers under which conditions these assumptions are most likely to be true.
Goldfarb says his paper should encourage future studies to examine the differences in the role certification contests play in industries with objective versus subjective determinations of quality: Whereas auto racing is conducive for observing the quality of cars in terms of speed and propensity to break down, the same does not apply to other industries, like fashion, wine and art, in which quality is determined more subjectively.
Read more: Did victories in certification contests affect the survival of organizations in the American automobile industry during 1895–1912? A replication study is featured in Strategic Management Journal.
Media Relations Manager