Small High Schools Post Big Gains: 5 Questions with Gordon Berlin
A chemistry classroom at the Collegiate Institute of Math and Science, a small school located in the Bronx. (Photo courtesy New Visions for Public Schools)
Small public high schools could be a key part of fixing urban education systems long plagued by poor academic performance, dismal graduation rates, and rare instances of college enrollment, according to a new study.
Research released today shows that small public high schools in New York City, which serve mostly low-income and minority students, posted 9.4 percent increases in graduation rates among students who entered the 9th grade between 2005 and 2008 – to 71.6 percent. Public school students in a control group yielded just a 62.2 percent graduation rate. Small high schools also boosted college enrollment among graduates by 8.4 percent – to 49 percent. The multiyear study was conducted by MDRC, a highly-regarded nonprofit, nonpartisan education and social policy research firm that aims to improve the effectiveness and efficiency of social programs by evaluating and demonstrating new and existing approaches and solutions.
Students in New York City public schools are matched to high schools through a combined application-lottery process. More than 120 academically nonselective small public schools were created in 2002 to cater to the diverse needs of city students. The schools were developed and approved by the city through a competitive proposal process that sought emphasis on academic rigor, student-faculty relationships and community partnerships. The movement was led in part by education nonprofit New Visions for Public Schools and the College Board, among others, to spur innovation in education. Funders included the Bill & Melinda Gates Foundation and the Carnegie Corporation of New York.
The MDRC study examines more than 12,000 students across the city’s traditional-sized high schools and 88 of the small schools of choice. RealClearEducation spoke with MDRC President Gordon Berlin about implications of the new findings. Here’s an edited version of our conversation.
What are the most important takeaways from the study?
The students enrolled in their small school of choice were 15 percent more likely to graduate within four years and 22 percent more likely to enroll in college – that’s despite the fact that 3 out of every 4 of these students were behind grade level when they entered the 9th grade. And yet the bulk of this increase in high school graduation rates is accounted for by students attaining Regents diplomas.
I feel like it’s the holy grail of the current movement in high school reform. We’re trying to prepare students for college and career, this is the first evidence at scale that a system was able to do this, where a high school reform led not only to increases in graduation rates but also to increases in college readiness, and then to college enrollment.
We’ve been able to follow one group of students now for several additional years, and this difference persists – they’re still in college. So this gives us some encouragement that we’re eventually going to see a difference in college graduation as well.
This is not the precious story of one outstanding high school of a few hundred students. This is the story of over 100 high schools in the city of New York, serving more than 40,000 students, and they were able to produce these results year after year. These students overwhelmingly live in the poorest communities in the United States, and 93 percent were students of color, 83 percent were low income, to get these kinds of results – the graduation rate for African American males increased by 22 percent, and the college-going rate is 36 percent higher than the control group.
How do your findings comport with other research?
I don’t think there’s anything comparable at this scale, with this level of reliability, with regard to high school reform. You’ll hear stories about a single high school with a few hundred students that produced positive results, but often those studies are not very reliable. That’s not the case in this study.
The study is unique because of the scale, the number of schools and students involved. It’s unique because of the unusually reliable lottery system that we were able to rely on for the research design. And I think it’s unique because of the results. You just do not see effects of this size.
Two things account for these effects: One is the inclusive, bottom-up process involving teachers, principals, unions, the Board of Education, community representatives, foundations. That process has not generally been followed when starting a new school – it tends to be top-down by local school district officials. If you replicated this process, you might do better and even replicate these results. A second reason is the characteristics of the schools themselves: they’re not just small, I can’t emphasize that enough. They use an academically rigorous curriculum coupled with high expectations, they were organized around a theme – for example, the school for social justice, sponsored by the largest law firm in the city -- and they use community and industry partnerships to increase their relevance. All students had a relationship on a regular basis with an adult – usually a teacher – and the adults set the culture and worked with each student.
It’s also important to recognize that most urban high school reforms are focused on school turnaround. For the most part, there is very little evidence, except in anecdotal evidence, that school turnaround has been effective. Just look at the federal results of the School Improvement Grants – their own data shows that there’s almost no effect.
Why is there such a pervasive sense that small schools didn't work when evidence like this keeps mounting?
This is one of those oft-told stories that turned out not to be true. An evaluation of small high schools was done a while back that didn’t have the ability to create a reliable counterfactual, a reliable comparison that would tell you what would happen if these schools didn’t exist. That concluded that these schools weren’t making much of a difference. I don’t know for sure about the other cities that it looked at, but it turned out to be wrong for New York City.
What are some caveats to these findings? Things we still need to be wary of?
This is an extraordinary foundation to build upon, but we still have a long ways to go. Only half the students went to college, a third didn’t graduate in four years from high school, we still have way too many entering 9th grade way behind when they arrive. And while these schools help to catch them up, the students who were the furthest behind didn’t improve enough to go on to college for the most part.
Another issue with small high schools is whether they can offer the diverse experiences, like AP courses, music, dance, and sports teams. There are solutions where you can get groups of schools to band together. I don’t think it’s necessarily clear that a small high school is the answer for every problem we face in the urban school system, so we recognize that there’s still more to be done, but I’d rather build on something I know is really making a difference.
How would you characterize the effectiveness of small schools more generally?
I think people need to pay attention to these results and realize the large high schools we created in another era shouldn't be the sole strategy for our urban high schools. I’m not saying we shouldn’t have any of them, but I think students need to have a variety of choices, that smaller schools enable the changes where they’re small not just in size but in function, they’re academically rigorous, provide the support that students need to make it. There just are not results like this and we ought to do more of what we know.
One of the reasons we haven’t done more is because urban districts are starved for cash and resources, and the very large high schools have economies of scale that are substantial. But if you look closely at these results compared to the students who went to control schools, these small schools have either similar or higher cost per pupil, but have a 16 percent lower cost per graduate because more students graduate in four years, and fewer need a fifth year of high school.