Dr. Henry M. Levin,
Co-Director of Center for Benefit-Cost Studies in Education, Columbia University (co-authored with Clive Belfield,  Queens College, and Co-Director of Center for Benefit-Cost Studies in Education, Columbia University)

Annually, 10,000 students leave Minnesota’s schools without a high school diploma. When measured over the lifetime the economic consequences if a dropout becomes a graduate are significant. We can express these consequences as the equivalent of a certificate of deposit in terms of their values at age 20. The individual graduate gains the equivalent of a CD worth $475,900 in extra earnings. The taxpayer gains the equivalent of a CD worth $251,900 in increases in tax revenues and in lower expenditures on health, crime, and welfare. The state of Minnesota gains the equivalent of a CD worth $1,059,500 from the individual and taxpayer benefits plus lower crime victimization and faster economic growth. The economic case for raising the high school graduation rate is therefore strong.  The moral case for raising the high school graduation rate is also powerful. High school graduation rates in Minnesota are stratified by race, sex, and family background. The disparities begin in early childhood but are magnified through the K-12 years. Many of the graduates then go on to college, leaving the high school dropouts further behind. Although Minnesota spends more on public education for disadvantaged students, the difference is not substantial. The search for effective educational interventions in K-12 schooling should therefore be intensive. The quest should be primarily for interventions to raise the graduation rate. But there is a strong correlation between achievement and graduation: our review estimates that an increase in 8th grade achievement of one standard deviation is associated with a 48% lower probability of dropping out of high school. So, the search should also consider reforms that increase achievement. However, the challenges to finding and selecting reforms are significant. Cost-effectiveness studies require that the alternatives have similar educational goals and comparable measurement of outcomes. Many studies claim to be educationally effective, but they are often based upon different measures of outcomes. Even when based upon test scores, the results may be for different subjects, different grade levels, different populations, and different test domains. Few studies that make claims for positive results meet even the minimal standards for rigor in evaluation design and implementation.  And, almost none provide useful information for measuring costs. These conclusions hold even if we expand our review to include interventions that raise achievement. After an extensive review, we found only a few K-12 education reforms that have demonstrated effectiveness in raising the graduation rate and whose costs can be calculated with reasonable confidence. These reforms are:     Unit costs per student    Extra high school graduates per 100 students    Costs per additional graduate Increasing teacher salaries     $2,850    5    $56,850 Reducing class sizes in elementary school across all students     $12,840    11    $116,720 for free lunch eligible students only     $12,840    18    $71,330 Success for All Elementary school reform    $3,842    4    $96,050 First Things First High school reform     $5,440    16    $33,680 Talent Development High school reform with career academy model    $2,790    8     $34,850 Check & Connect High school mentoring and monitoring program    $8,150    17    $47,930 Achievement for Latinos through Academic Success High school program monitoring behavior and academic success    $3,940    5    $78,860   When we compare the costs of these reforms to the economic benefits that flow from high school graduation, we find the benefits to the taxpayer easily exceed the costs:     Benefits to the taxpayer divided by the costs of the intervention Increasing teacher salaries     4.01 Reducing class sizes across all students     1.96 Reducing class sizes for free lunch students     3.21 Success for All    2.38 First Things First     6.72 Talent Development    6.56 Check & Connect    4.77 Ach. for Latinos through Academic Success    2.90 There are many other interventions with potential benefits, but either their effectiveness is not fully demonstrated or there is inadequate information on costs. Effective programs that may pass a cost-benefit test, but presently lack rigorous evaluation data and cost information, include: •    Mentoring programs (12 Together, I Have a Dream) •    College readiness programs (CAP, TEACH, Talent Search) •    Employment readiness programs (Career Academies) •    Extended hours programs (After-school, summer school) There are other areas of reform where more knowledge is critically needed for reforming K-12 education. These include: specific programs (such as AVID, Project GRAD, KIPPS academies, ISA model); reform related to teacher quality; and the range of family interventions. (Pre-school is outside the scope of our review).  Overall, the evidence on what works – and what interventions pass a benefit-cost test – is very limited. Accurate and detailed information on resource requirements necessary to estimate costs is as sparse as rigorous evaluations on effectiveness. A major challenge is to fill these crucial gaps to provide a more comprehensive list of educational interventions by their cost-effectiveness and cost-benefit results. One method developed in this paper to expand the numbers of programs that predict increases in high school graduation is to provide a framework for converting test score increases at 8th grade or above into impacts on high school completion. This technique will enable us to embrace a wide range of high school reforms that show improvements in student achievement, but do not follow up students to graduation to show the impact on high school completion.

Click here to read the full paper.

Powerpoint available here.