An experiment in incentivizing innovation

One of the animating spirits of the rise of STEM education is the push for innovation—new technologies, new applications, new solutions to intractable problems. But is cultivating that creative ability as common an outcome in students as tech enthusiasts would lead us to believe? A recent study by a team of researchers from the University of California San Diego (UCSD) attempted to determine whether a gift for innovative thinking is merely something that prompts students to choose STEM classes, or whether it can be cultivated among those who believe they do not have that gift. The conclusion is that with billions of dollars of investment in STEM, American K–12 education could be putting its eggs in an unstable basket.

The locus of their work was an app-design contest open to all undergraduate students taking classes at UCSD’s Jacobs School of Engineering, whose winners would be determined by a panel of tech entrepreneurs and executives. Prize money was offered for the top three finishers, and the winners need not have progressed as far as a useable electronic prototype in order to compete; judges would be looking at things such as written plans and design mock-ups as seriously as they would live products. Motivated by outreach such as newsletters, emails, and information sessions, 103 students applied to participate in the contest as “self-selecting innovators.” A random group of students who did not sign up for the contest was then offered a $100 financial incentive to sign up. Eighty-seven students did so, and the researchers categorized them as “induced innovators.” These two groups were then randomly split with half of each group given a further treatment in the form of confidence-boosting emails (a “motivational treatment”) through the first several weeks of the contest. Data collected included pre-contest and post-contest surveys and the amount and quality of contest submissions.

Predictably, the majority of self-selecting participants were engineering and computer science majors, while the induced innovators were less likely to be drawn from these fields of study. Additionally, the induced innovators had lower GPAs than their self-selecting peers.

The results: The financial inducement seemed to achieve the goal of expanding the number and diversity of contest entrants. More importantly, the two groups ultimately submitted projects at more or less the same rate despite the difference in assumed capability and motivation for innovation. The submission rate for the full cohort of participants—approximately 10 percent—was deemed normal for such contests. The motivational treatment, however, showed an insignificant impact on both the number of projects submitted and the quality of those projects overall, although the small submission numbers made definitive analysis difficult. In the end, average judges’ scores were indistinguishable between self-selecting and induced innovators; the quality of their work outputs were essentially equal.

While the evidence suggests that “innovativeness” is not strictly an innate trait of certain individuals, the effects of financial inducement and motivational treatment interacted in surprising ways and had some counterintuitive effects on student performance. For example, though the induced innovators with below-median GPA performed the worst of all the groups, they were the individuals who benefitted most from the motivational treatment. Motivational treatment led to a small boost in project submission rates and project quality ratings. Additionally, those individuals receiving motivational treatment who did not ultimately submit a project were more likely to report that lack of time was the reason rather than perceived lack of ability to compete. The highest performers in the contest overall were self-selecting innovators also with below-median GPA, but the motivational treatment appeared to exert downward pressure on their average scores by comparison to the above-median GPA self-selecting innovators. The researchers concluded by raising questions related to individual characteristics not tested. What stops a student who can succeed in the innovation task from self-selecting into the contest? How much financial inducement is enough to incent innovation? Why do motivational messages received by self-selecting students with the most apt background and qualifications appear to interfere with the highest levels of successful innovation?

Although this study examines a STEM initiative in a college setting, who participates in and who benefits from the boom in K–12 STEM education is also a vital question. Future English majors, future attorneys, and many other students could be left behind if technology education becomes the pre-eminent focus of K–12. Additionally, STEM education without innovation could easily become a muddled mess of poorly coordinated traditional lesson plans cloaked in twenty-first-century buzzwords. This new research asserts that the pipeline of innovators is larger than it would at first appear. Proper incentives along the way seem to be a promising way to bring more students into the STEM fields, but we also need to know how to help them be successful once they’re in.

SOURCE: Joshua S. Graff Zivin and Elizabeth Lyons, “Can Innovators be Created? Experimental Evidence from an Innovation Contest,” National Bureau of Economic Research (February 2018).

Jeff Murray
Jeff Murray is the Ohio Operations Manager of the Thomas B. Fordham Institute,