American high school students score higher on low-stakes tests and are more willing to answer test questions when they are motivated by cash, according to new economic research.
Giving high school-aged students money as a reward for performing well improved scores by roughly 5 percent, according to economists at the University of California, San Diego and the University of Chicago.
This outside incentive, they concluded, suggests that the education gap between U.S. students and their international peers may have less to do with understanding the material and more to do with effort.
Critics of the U.S.'s educational system often point to the surveys done by the Organization for Economic Co-operation and Development's Programme for International Student Assessment, which consistently rank American students well below their peers from other major economies.
Seventy percent of American students graduate from high school, putting the United States in the bottom quartile of OECD countries. Among the 65 countries that participated in the OECD"s 2012 assessment, which is the one the researchers used, U.S. high school students ranked 36th for mathematics.
But using money as a motivator, the researchers discovered, could lift the U.S. to 19th in the math test rankings.
In their research, the economists compared Shanghai students, who are top performers on international assessments, to Americans. They created a 25-question, 25-minute mathematics test consisting of previous PISA questions.
On the day of the test, some of the students in each experiment group were given $25 in cash (or an equivalent in reenminbi for the students in China) and told the money was theirs to keep but the researchers would take away $1 for every wrong or incomplete answer. Some of the students were not offered money.
While the performance of Shanghai students didn't change, the American students who were offered money attempted more questions and were more likely to answer those questions correctly, the economists found.
"We estimate that increasing student effort on the test itself would improve U.S. mathematics performance by 22 to 24 points [on PISA], equivalent to moving the U.S. from 36th to 19th in the 2012 international mathematics rankings," the researchers said.
Source: Sadoff et al., 2017.
They also found that students offered money scored better in the latter half of the test, with scores increasing three percentage points in the first half of the test and eight percentage points in the second half.
The probability an American student answers a question correctly typically drops six percentage points if the question appears at the end of the test, suggesting motivation may be at work.
"The general insight is that maybe students aren't trying as hard as they could be," Sally Sadoff, an economist at U.C.S.D., told CNBC in an interview. "It takes mental effort, mental power."
Of course, Sadoff said, American students perform better when the results of effort are readily apparent -- on high-stakes SAT tests or final exams, which have significant effects on college admissions or grade point averages. But the fact that many students lose academic drive outside of these select exams may betray a typical attitude toward schoolwork.
Money incentives increased the score of American male students by nearly twice as much as female students, the research also found, which confirmed existing literature on gender differences.
Sadoff said the goal of the economists' latest paper was not to study how incentives work, "but rather to use incentives as an experimental tool to understand the interaction of culture with motivation to do well on the test."
Source: Sadoff et al., 2017
To be sure, Sadoff's group aren't the first economists offer students – or teachers – cash for improved performance. Multiple studies have shown that cash payments do tend to improve performance on individual assignments, but those effects tend not to last longer than the duration of the studies.
Harvard's Roland Fryer has spearheaded much of the research effort looking at student motivation in the past decade.
In a 2012 paper that appeared in The Quarterly Journal of Economics, Fryer spent $6.3 million on a study that found offering students money for standardized test grades doesn't work.
In this research, Fryer paid New York City students for performance on a series of interim exams; in Chicago he paid ninth graders every five weeks for solid core grades; in Dallas he paid second graders $2 per every book read; and in the District of Columbia he paid students for attendance and other positive behaviors.
As it turns out, incentives for grades or scores had little effect, Fryer found. Instead, paying kids to read had a statistically significant effect on comprehension, vocabulary and language skills.
Some may raise ethical objections to monetary incentives to improve effort on low-stakes tests, but Sadoff said the results of the research do raise an interesting question.
"We've been working on studies on motivation for a while," she said. "I think for a lot of kids, their parents are just instilling a habit [of good study practices]. But for some kids, I think there is a role for us. If the return to education is so high, why aren't kids trying?"
Note: Working papers have not been peer-reviewed or been subject to review by the NBER Board of Directors that accompanies official NBER publications.