The question, in a call last spring from the White House, seemed innocent enough: Would the Navy be interested in funding an expansion of a private-sector program encouraging high school students to take more rigorous math and science courses into schools that serve large numbers of military dependents? Its support would help promote a new, high-profile White House effort to help the nation鈥檚 armed forces and their families. But in agreeing to participate, the Navy also put itself ahead of the curve on federal education policy by embracing the untested and controversial idea of paying students who do well on standardized tests.
Cash incentives to students and their teachers are a core element of the 4-year-old, Texas-based , or NMSI, which targets low-achieving and low-income high schools with large minority populations. NMSI leaders believe that those schools can improve the quality of instruction by offering more Advanced Placement courses, on the assumption that their students can handle more challenging material if given the chance. The incentives are part of a well-structured program that also features extra lab equipment, Saturday classes, and special training for teachers.
The Obama administration has no official position on the use of cash incentives, and Congress has never addressed the topic in legislation. At the same time, U.S. Secretary of Education Arne Duncan has voiced support for the concept as a tool for raising student achievement, and a small departmental program to foster participation in AP courses allows officials to give money directly to students. However, only two of the 55 current grantees are doing so.
STEM Investment
Despite the fact that it would be entering uncharted waters, the Navy decided to proceed. Reaching schools that serve a large number of military families is a good way to address the Navy鈥檚 concern about finding enough U.S. citizens to fill the science- and technology-based jobs required to wage modern warfare, said Michael Kassner, the director of research at the Office of Naval Research and head of the Navy鈥檚 STEM鈥攕cience, technology, engineering, and mathematics鈥攐ffice. 鈥淲e know that students whose parents are in the military are more likely to go into the military,鈥� he noted.
Mr. Kassner also had the money. The Navy is in the midst of doubling, to more than $100 million, its investment over five years in STEM education at all levels. 鈥淭he AP test appears to be successful in improving a student鈥檚 skills,鈥� said Mr. Kassner, a former professor of aerospace engineering at the University of Southern California. 鈥淎nd we thought NMSI looked like a good way to go.鈥�
Mr. Kassner agreed to the White House鈥檚 offer to sign onto a program, the Initiative for Military Families, which brings the NMSI model to schools serving a large number of military families. The White House effort is part of Joining Forces, a program launched in April that received a national shout-out from first lady Michelle Obama before the start of this year鈥檚 first World Series game in St. Louis.
NMSI officials compiled a list of 154 schools around the country near military installations and screened them for their willingness to follow NMSI鈥檚 formula. At the end of September, the Office of Naval Research a $1.1 million grant to NMSI to conduct a three-year pilot at three public schools in Virginia and Hawaii. Mr. Kassner says the Navy is open to funding many more schools if the initiative shows it鈥檚 capable of priming the STEM pump to meet the Navy鈥檚 needs.
Missing the Target?
At one level, NMSI鈥檚 approach appears to be working well: In the midst of a nationwide surge in AP test-taking, the number of minority and female students passing those tests at 228 NMSI-sponsored schools this year increased four to 10 times faster than for the country as a whole. Apart from the initiative鈥檚 own data, however, there is scant evidence in the research literature that the program can deliver the talent sought by the Navy and a host of other employers.
Kirabo Jackson, a labor economist at Northwestern University in Evanston, Ill., has looked at the Texas precursor to NMSI. His 2008 study is perhaps the only one to take a comprehensive look at the role of incentives in the targeted population, although it doesn鈥檛 measure the impact of incentives apart from the other program elements.
Mr. Jackson found that the strategy increased AP participation rates and boosted the number of students with high scores on national college-entrance tests. At the same time, he found the program didn鈥檛 increase high school graduation rates or the number of students taking college-entrance exams. That result, he said, suggests it鈥檚 more likely to help high-achievers already headed to college than to raise the aspirations of those who hadn鈥檛 planned to continue their education.
The jury is still out, Mr. Jackson said, on whether efforts to promote AP courses with cash incentives actually result in the kind of payoff that the Navy is hoping for, namely, more students with college degrees in STEM fields from high-quality institutions. Still, his overall assessment of the program is positive.
鈥淚t鈥檚 one of the few programs that does something good for these students,鈥� Mr. Jackson said. 鈥淢ost programs haven鈥檛 been evaluated rigorously. And I don鈥檛 know if it can be expanded to other settings, with other populations. But if a school district had $1 million to spend, I think a program like this is a good investment.鈥�
A 2010 by economist Roland G. Fryer Jr. of Harvard University delivers a much more sobering message about the value of cash incentives for younger students. Mr. Fryer conducted a randomized trial of experimental programs in four large urban districts involving 38,000 children in grades 2 to 9. Although the program elements varied greatly from one district to the next, he found that paying for outputs, such as test results, didn鈥檛 work and in some cases resulted in lower scores. On the other hand, he found that paying students for inputs鈥攕howing up for class, staying on task, reading a certain number of books鈥攈ad a positive effect.
Mr. Fryer speculated that students who don鈥檛 understand, for example, how hard work translates into better test scores are less likely to be motivated by the promise of cash rewards at the end of the term than by immediate reinforcements. 鈥淪tudent incentives based on inputs produce similar gains in achievement at lower costs,鈥� he concludes in a paper posted by the National Bureau of Economic Research.
An Education Department official who requested anonymity acknowledges that paying students is controversial. 鈥淧oor kids don鈥檛 always know how to do well on tests by taking the time to study and by learning how to take a test,鈥� the official notes. The official also admits that the NMSI approach assumes that 鈥減assing an AP test is a good thing鈥攚hich is a big assumption.鈥�
NMSI officials recognize that the program sets a high bar that may be difficult for some schools to clear. A school in which five students passed AP biology last year, for example, would be expected to record 37 passing scores this year, said Gregg Fischer, the director of the initiative鈥檚 AP training and incentive program. If its success rate was 16 percent last year, he said, that would mean enrolling some 200 students this year.
鈥楲ukewarm鈥� to Cash
At Green Run High School in Virginia Beach, Va., one of the schools getting Navy funding to pursue the NMSI program, many students 鈥渃ome in without all of the tools they need to succeed in an AP course,鈥� said Principal George Parker. As a result, the school鈥檚 success rate to date on the end-of-year standardized tests has been in the single digits for some AP math and science courses. About one-fourth of the school鈥檚 1,640 students come from the military community, which has a high student-turnover rate. The Joining Forces initiative hopes to provide a seamless transition for those students when their families relocate. (鈥淪chools May Be Asked to Report on Progress of Military Children,鈥� May 11, 2011.)
Students receive $100 for each math, science, and English AP test on which they score at least a 3, on a 5-point scale, and teachers earn $100 for every successful student, with the possibility of bonuses up to $3,000 for increasing AP participation, but Mr. Parker said money isn鈥檛 the real objective.
鈥淭he rigor of taking an AP course better prepares students for college, even if the students don鈥檛 pass the final exam,鈥� he said. Many students are so far 鈥渓ukewarm鈥� toward the program, he added, because of the extra work required to be eligible for the payments. And Mr. Parker doesn鈥檛 think that the bonuses provide any additional motivation for his teachers to do their jobs well.
In many states, teachers鈥� unions are adamantly opposed to cash incentives, In addition to clashing with most labor agreements, incentives are viewed as undermining the learning process. In New York City, for example, the United Federation of Teachers declined to participate in that component of a program serving 31 schools that is otherwise modeled after the NMSI efforts. And in October, officials for the local nonprofit that runs the program, REACH, for Rewarding Achievement, decided to drop the student incentives as well when faced with a budget squeeze.
鈥淲e found that it prompted more students to take AP courses, but we didn鈥檛 see the magnitude effect that we had hoped,鈥� said Kathrine Mott, REACH鈥檚 executive director. However, Ms. Mott said that her organization will continue to offer professional development for teachers, Saturday classes, and classroom grants.
