The impending perjury trials of baseball greats Roger Clemens and Barry Bonds may seem far removed from science. But economists tell us that the world of Major League Baseball bears a strong resemblance, in structural terms at least, to that of academic scientists. Professional athletes who use performance-enhancing drugs can, it turns out, throw light on scandals involving prominent scientists who enhance their results, such as Woo-Suk Hwang, formerly of Seoul National University, and, more recently (and allegedly) Marc Hauser of Harvard University.
Like every such revelation, Harvard's August disclosure of Hauser's alleged scientific misconduct sparked interest in the causes and frequency of scientific deception. Some people seek answers in personal psychology or cultural values. But sociologist Andrew Schrank of the University of New Mexico, Albuquerque, tells Science Careers, "The core point is that it's about the structure of the organization."
What can structural analysis tell about scientific misconduct? First, along with professions such as big-time sports and entertainment, academic science is what economists call a tournament field, which makes it susceptible to cheating. And second, according to a study by Schrank and fellow sociologist Cheol-Sung Lee of the University of Chicago in Illinois, social structure explains why it happens much more often in some places than in others and what types of arrangements can discourage it.
Distribution of deception
Some professions -- high school teaching or family dentistry are two of many examples -- lack opportunities for stardom and distribute rewards fairly evenly among participants. Tournament fields, on the other hand, give small numbers of people very big prizes. Very few athletes and actors win riches and stardom; a tiny percentage of scientists win "a named chair, scientific renown, awards," write Harvard economist Richard Freeman and co-authors. "Amplifying small differences in [performance] into large differences in recognition and reward ... [creates] intense competition," they add. All that separates celebrities from the host of others may be hitting a couple of home runs, landing a certain lucky part, or being first to publish a significant result. Cheating can, if undetected, provide that career-making edge.
What determines whether scientific dishonesty happens frequently or rarely is the structure of the institutions where scientists work, argue Schrank and Lee in their article titled "Incubating Innovation or Cultivating Corruption?: The Developmental State and the Life Sciences in Asia," which was published in March in the journal Social Forces. The article compares the situation of life scientists competing at the cutting edge in the United States and several Asian countries. It's most useful to think "not about the individual cases but about the broad distribution" of deception, Schrank says. Although the true extent of wrongdoing cannot be known and many miscreants escape detection, "there's every reason to believe that the broad distribution ... in the United States is far less tainted than the broad distribution in Asia." According to studies reported in The New York Times in October, for example, half of the scientists queried in one survey in China claimed to know someone who had committed academic misconduct. In another survey, a third of scientists at top Chinese institutions said they had committed misconduct themselves.
The German model
Universities in Northeast Asia are "particularly prone to scientific scandal" because of a specific combination of structural factors, argue Schrank and Lee in their article. One structural factor is the enormous rewards that governments bestow on conspicuously successful senior scientists as incentives "to overtake their advanced industrial counterparts." The "illiberal laboratory cultures inherited from Germany by way of Japan" is another.
Universities in South Korea, Japan, and China retain the highly authoritarian 19th century German organizational model, Schrank says. In Germany itself, that traditional system remains "alive -- if not necessarily well" and "a substantial body of literature" indicates a "link between illiberal labs and scientific fraud" in that country, Schrank and Lee write. "A spate of research misconduct shocked not only Germany but the whole scientific world" in the 1990s, they continue. "In the most notorious case, two German oncologists were convicted of fabricating or falsifying data in more than 50 articles." Since then, reforms have "deemphasize[d] citation counts and empower[ed] ombudspersons to protect whistleblowers," Schrank and Lee note, but the reforms "are new, and therefore unproven."
Numerous "high profile" cases have emerged in South Korea, Japan, and China, where some people "worry that disclosure is the exception to the rule, especially among the scientific elite," Schrank and Lee state. In the interview, Schrank hastens to acknowledge that American scientists also cheat, but, he adds, "We simply think that the incentive structure and the organizational structure in Northeast Asia are much more conducive to this type of fraud than either the incentive structure or the organizational structure in the United States."
"There is only one full professor [in each department] in the classic German chair model," Schrank explains in the interview. "There might be associate and assistant [professors] and lecturers and senior scientists, ... but not as diffuse a hierarchy as you would have in the United States."
American departments have multiple full professors -- numbering dozens at large universities -- and a chair who is only a first among equals. This multiplicity of senior faculty members means that no single person rules a department. It also provides the scientists lower on the totem pole -- grad students, postdocs, staff scientists, and others -- the possibility of moving among labs and mentors; their careers need not be tied to a single sponsor. By contrast, "there's no lateral labor market in northeast Asian universities," Schrank continues. "You really are dependent upon the lab director for your entire career prospects."
Incentives for deceit
The tremendous prestige and power, plus the major money, that Asian governments award professors who raise the national profile by scoring marquee publications create incentives to enhance results. American scientists, of course, also feel intense pressure to publish -- but, Schrank and Lee argue, the structure of U.S. laboratories makes getting away with fraud much harder and researchers therefore much less likely to risk their careers by trying.
In German-model labs, Schrank and Lee write, scientific authority and the supervision of work flow in only one direction: downward from the all-powerful full professor. Noting research by sociologist Jason Owen-Smith, who is now at the University of Michigan, Ann Arbor, they argue that American-model labs organize work not around a singular authority figure but around formalized "skeptical encounters." Regularly scheduled staff meetings, a prominent feature of American lab life, provide opportunities for scientists at all levels to question, openly and publicly, their colleagues' -- and even their superiors' -- data and conclusions.
Doing so is, in fact, a crucial means of career advancement. "If a junior person challenges the senior person and is successful, his or her status rises enormously, so there is an incentive to do it," Schrank notes in the interview. By contrast, says "a Harvard-trained Korean biologist" quoted in the article, "frank evaluations of colleagues are simply 'not part of the scientific culture' in his homeland."
So how do you lessen the likelihood of fraud? "I think the way to combat this is to recognize that it's not Asian culture, but that it's organizational, and then to think about what types of organizations militate against it," Schrank says. He cites as evidence Taiwan and Singapore, which share an Asian cultural background with China, South Korea, and Japan and "have pursued the same sort of intensive drive to give incentive for blockbuster publications, with monetary rewards," but haven't seen nearly as much scandal.
The reason? They "have broken with the German model. Taiwan is particularly interesting" because it used to follow the highly hierarchical German system, but it changed to the American one. Singapore has consistently followed a British system that is also similar to that of the United States.
The key is institutional structures that value, encourage, and reward skeptical interchange among personnel at all levels. Events such as a regularly scheduled lab meeting can discourage would-be cheaters "precisely because every week they have this confrontation and someone can call them on it." The research systems in more scandal-prone countries "don't have this sort of built-in mechanism for quality control, and as a result there isn't a structured space for that sort of challenge."
Although Lee and Schrank hold up the U.S. system as one whose structure encourages good conduct in research, "we think there's probably too much hierarchy in U.S. labs as well," Schrank says. The Hauser case, in which it was reportedly junior members of the lab who unmasked wrongdoing despite Hauser's efforts to dissuade them, is an example of how hierarchy and integrity are often in competition.