For me, cage-busting is a way of life... unfortunately, according to my wife. When Dave Levin and I created the first KIPP program in Houston with 47 fifth grade students at Garcia Elementary School in 1994, we were cage-busting within the system by extending the school day and putting our underserved students on a different life trajectory. Since then, KIPP has grown to a national network of 125 public charters schools in 20 states serving 41,000 students, 85 percent of them are from low-income families. Today I want to share some brand-new research that’s helping KIPP assess how we are delivering on our promises to children and families.

At KIPP, we firmly believe that what you measure matters. It’s not enough to break out of the cage; you also have to show that what you’re doing outside the cage is working. KIPP has always welcomed independent research into our model and methods. But five years ago, we took an even bigger step into the world of data and analysis.

In 2008, Mathematica embarked on a multi-year study of KIPP middle schools nationwide. Our rationale for green-lighting this study was simple: we wanted to understand what was really going on in our schools, and share it transparently with the public. We chose to work with Mathematica because we knew they would dig deeply into all of KIPP’s results--the good, the bad, and the ugly.

Now, in a new report--just released this morning--Mathematica has dug deeper into KIPP’s results, looking at achievement results across 43 schools and four academic subjects, using new methodologies and research tools. Mathematica found that:

- KIPP produces achievement gains across all middle school grade levels and academic subjects. Mathematica looked at nearly double the number of schools as they did in their 2010 report, and double the subjects. They found that KIPP’s gains are consistent across all years of middle school, and across every subject studied.

- KIPP’s academic results are substantial. Not only are KIPP’s students making gains in middle school math, reading, science, and social studies, but the gains they’re making are large enough to significantly close race and income based achievement gaps.

- The matched comparison design produces estimates of KIPP’s achievement impacts similar to estimates of the same impacts based on an experimental, lottery-based design. The lottery analysis was able to account for factors like student and parent motivation, since every student in the sample was equally motivated to sign up for the lottery.

- KIPP’s student gains hold up on low stakes national tests. The academic gains KIPP kids showed on the state tests were repeated in the national norm-referenced tests, meaning that KIPP is not “teaching to the test.”

-KIPP schools are not creaming the most able students. When it comes to prior achievement, KIPP students tend to enter 5th grade with test scores that are lower than the average for neighboring elementary schools.

Mathematica is also providing us with crucial information about where we need to improve. The researchers reconfirmed that we are enrolling lower numbers of students receiving Special Education services and students who have limited English proficiency than district schools - not dramatically lower to alter the findings above, but one of our key internal questions at KIPP is: are we serving the children who need us the most? Any data points where we are even slightly under is something for us to flag. Mathematica’s drill down into our achievement data also showed that our students’ gains are higher in math and science than they are in reading or social studies. These are two issues we can now focus on with clarity, and apply our spirit of cage-busting to try and find new solutions.

With all that said, the part of this report that has the most fodder for cage-busting leaders is towards the end, where Mathematica looks at the factors that make schools successful.

Mathematica noticed that some KIPP schools produced bigger academic gains than others. They identified two characteristics common in our most successful schools. The first is leaders who set strong, cohesive behavior standards, so that students know what is expected of them. The second is time spent learning. Although all KIPP schools have extended days, some have longer days than others; the KIPP middle schools with the largest impacts are not those with the longest school days, but instead are schools that spent relatively more time on core academic subjects.

There are lessons here not just for KIPP, but also for our district counterparts in the communities we serve. School districts can adopt elements of our most successful schools by giving principals more autonomy to establish healthy school cultures and to expand the school day to include more time for learning math and reading without having to sacrifice the other core subjects or the co-curricular classes. School districts in the Houston area are starting to do just that, as they realize that life outside the cage is preferable. Houston ISD has taken a competition-at-all-costs approach and has lengthened the school day and year in their chronically failing schools to try to emulate what KIPP, YES, and other cage-busting, high performing schools are accomplishing. Other Houston districts, like Spring Branch, Aldine, North Forest, and Galveston, have taken cage-busting one step further and are partnering with the cage-busters to start schools together and so they can start cage-busting themselves.

We will have really reached a cage-busting milestone when we give all public schools - charter and district schools - the freedom to set children up for success in college and in life, and when cage-busting will be a confusing term, given all that remains is some scrap metal from the old days.

-- Mike Feinberg