In the federal legislation signed this spring to reform student lending, one feature has been largely overlooked by the press: The new law increases the incentive for college graduates to choose public-service careers, such as teaching. But in a time of high unemployment and spiraling deficits, an important question is whether there is any evidence that public-service subsidies help bring the best and the brightest graduates into the jobs where they are most needed.
The premise underlying government financial aid for people who take public-service jobs is that the low pay and stressful working conditions in many of these jobs—think military service or teaching in a high-need school—do not match the public benefit the jobs provide. The theory is that financial incentives will help attract skilled professionals to such jobs. But there is surprisingly little evidence to show whether the theory is correct. If, in fact, the financial incentives go to college graduates who would have taken the jobs anyway, they are not a good use of scarce funds in an era of record deficits.
A recent study we wrote, which appears in the Summer 2010 issue of the Journal of Policy Analysis and Management, tried to answer this question. We found that certain types of public-service incentives can indeed help high-need professions attract skilled workers.
We assessed the impact of a $20,000 fellowship aimed at attracting academically talented teachers to designated low-performing public schools in California. This award, called the Governor’s Teaching Fellowship, was granted competitively to about 1,200 students in graduate-level teacher education programs during the 2000-01 and 2001-02 academic years. Recipients were required to teach in low-performing public schools for four years after earning licenses, or repay the fellowship as if it were a student loan. The annualized benefit amounted to about 15 percent of a new teacher’s salary in California.
To estimate the award’s impact on teachers’ behavior, we examined the first jobs of teachers with strong academic backgrounds who pursued teaching licenses before, during, and after the award was available. We used similar newly minted teachers to control for broader labor-market conditions. We found that receiving a fellowship increased a novice teacher’s probability of working in a low-performing school by 28 percentage points—a substantial effect.
This means that about two out of every seven teaching fellows who took positions in low-performing schools would not have done so without the fellowship. And three-quarters of recipients in the study who began working in low-performing schools remained in these schools for at least four years. We calculated that the state spent about $245 in bonuses per student taught by a teaching fellow who would not have otherwise worked in a low-performing school. This is roughly 3 percent of what the state was spending on the average student in California in 2001.
We should be cautious, of course, in generalizing these findings to other career fields and incentive plans. For instance, when studying the decisions of new attorneys to work in public-interest law, Harvard economist Erica Field found that their choices were affected not just by the benefit amount but by when the money was paid. Benefits paid during law school as fellowships—similar to the award we studied—were more influential than financially equivalent loan-forgiveness benefits paid after law school.
Incentives are good investments of scarce public funds if they not only influence workers’ behavior but also seek to correct well-understood market failures. For instance, it has long been known that low-income and lower-performing students are less likely than their peers to have teachers with strong academic backgrounds because of the differences in working conditions among schools. This is especially troubling because, all else being equal, teachers with stronger academic backgrounds (as measured by standardized-test scores and college competitiveness) are modestly more effective at raising student achievement. The fellowship we studied in California sought to mitigate this problem. But it is not clear that recent federal loan-forgiveness legislation will do so.
Historically, public-service education benefits have targeted workers in difficult jobs, such as teachers in low-income schools, health-care workers in isolated areas, police officers, and military-service members. Yet the Public Service Loan Forgiveness Program, which was passed by Congress in 2007 and became effective in 2009, defines public-service careers more broadly. Though subject to income caps, its definition includes not only police officers, military-service members, social workers, and public-interest lawyers, but also employment in virtually any other federal, state, or local government entity or nonprofit organization, including most colleges and universities.
The definition also includes public school teaching, but without regard for the academic performance or family incomes of students in the school. In other words, a teacher in an affluent public or a nonprofit private school is eligible for the same benefits as a teacher in a school that serves many disadvantaged students, though schools in the latter category typically have a harder time finding and keeping skilled teachers.
To be clear, the student-loan-reform legislation signed this spring will eventually make income-based repayment terms more generous for all eligible lower-income workers by reducing minimum monthly payments and forgiving loans in 20 years instead of 25. Yet workers whose jobs meet the broad definition of public service stand to reap by far the largest benefits, because their loans can be forgiven after only 10 years.
These terms may make college graduates more eager to take public-sector over private-sector jobs, but all public-sector jobs are not created equal, nor are they deserving of equal subsidies. In an era of necessary thrift, policymakers should aim education subsidies at bringing top talent to the nation’s most critical jobs, including its highest-need schools.
