Do Grades As Incentives Work?
Why "smart" incentives can never be smart enough.
Posted Jan 09, 2011
This post is co-authored by Barry Schwartz and Ken Sharpe.
We've written in earlier posts about the widely shared presumption that the way to get the people we rely on to do the right thing—the doctors, the lawyers, the teachers, the bankers—is by creating the right incentives. Smart incentives. This belief is pervasive in modern American society, but it is wrong. Consider an incentive system we know a lot about, with 80 years of experience between us—college grading.
There is a common presumption among many teachers that grades make students work harder and learn more. There is obviousness to this argument that seems to make it unquestionable. As a general matter, we know that rewards and punishments can change behavior. As teachers we see individual students work harder—get more serious—after their first bad grade on a test or paper. We imagine that guaranteeing all our students an A- in the course on the first day would lead to a dramatic decline in preparation, class discussion, and quality of writing. And if all the other courses they were taking that term were graded, and they had been brought up on system based on grades, that's likely to be true (one of us once tried it as an experiment, with dismal results). So we could amass considerable evidence that grades get students to buckle down and learn more, and that tougher grading would do this even better. Most faculty discussions of grading presume this truth. That's why discussions about grade inflation are framed as being too soft on students—too concerned about minimizing feelings of stress and maximizing feelings of empowerment and not concerned enough with actual achievement.
There is no question that we can use grades to get students to change their behavior, but are we getting them to learn more? One danger is that grade-focused teaching corrodes the very meaning of learning. The purpose of learning becomes merely the achievement of grades. Not the mastery of the material. Not finding innovative and imaginative solutions to tough problems. Not joining with fellow students to run with an idea and see how much each can learn from the others. It becomes instead what former Harvard dean Harry Lewis calls "an empty game of score maximization." It makes the work seem pointless. This is an old, if not always salient, concern. Lewis points to an 1885 Harvard Crimson article on the school's rigorous "Scale of Rank" system: grading to make students work "encourages an unscholar-like tendency to work for marks, and prevents the establishment of high motives for study. Students are dwarfed by it, to the low stature of grinds for marks. Injudicious selections of courses are encouraged by it. Cribbing thrives under it." Lewis argues that the gamesmanship encouraged by today's grading system encourages cheating because "students realize that the game they have to play is meaningless, and their commiseration emboldens them to dishonesty." If the only purpose of learning is getting the grade, the only reason not to cheat is the fear of being caught. That encourages an increased cat-and-mouse frenzy—a system of mutually assured escalation—as students use internet tools and programs to plagiarize others papers—or even hire other students to write them—and faculty turn to computer programs designed to catch them.
When grades-as-incentives encourage gamesmanship instead of excellence there is another way that this corrodes learning. A student out to maximize her grade point average is tempted to choose the easiest courses, those with the least challenge and work, or those with the "easy-grader" professors. Dean Briggs, at Harvard at the turn of the twentieth century, called such students "mark-fiends" who never come to anything later in life. Pressure mounts as graduation approaches and students vie for high rankings and honors. Do I want to risk the summa cum laude by taking that tough physics or philosophy course?
There is another pernicious downside to motivating with grade incentives. The students in the bottom half of the class—students whose learning we want to encourage—know that the odds of high grades and high rankings are stacked against them. If we corrupt students' souls by convincing them that the main motive for learning are high grades and honors, we end up de-motivating, and de-moralizing, those students who have little chance for the top rankings.
If grades corrode learning and de-moralize some students, what would happen if individual teachers started eliminating them—or standardized grade inflation by guaranteeing a common grade to the whole class on the first day? That would likely be disastrous. Students have grown up in a system that has taught them to work for grades. Most teachers are still using grades to incentivize students. The university's culture, ranking system, and credentialing depend on grades. This all creates extraordinary pressure on students. Even if they are ones who already know they thrive on their excitement and passion about the material and the skill and enthusiasm of a good teacher, when the time crunch comes, the pressure to put the time into the graded class is difficult to resist. Without a more systemic change, an individual professor would be asking a lot by assuming a change in her class would make a positive difference.
So do grades-as-incentives work? Students brought up in a system of incentives get accustomed to working for grades. So yes. It works for many students to motivate work. But if "working" means learning, these external incentives teach the students the wrong thing to aim at, the wrong reason for doing it, and often the wrong way to do it. If we are hoping our students will be life-long learners, why would they continue to learn in the grade-less post-graduation world?
That said, it is important to acknowledge that grades do serve an essential function. They provide feedback. "Did I get this right?" "Was my argument sound?" "Do I really understand differential equations?" We all learn from our mistakes, but we can't learn from mistakes that we don't know we're making. The trick is to focus on the feedback function of grades more ("Are my grades telling students what they need to know about their performance?") and the instrumental function of grades less. Doing this will take time and effort, but it will be rewarded by students who continue to learn even after they graduate—students who do their studies for the right reasons.