When families have the ability to choose schools--rather than be assigned to them based on their address--they'll have the opportunity to seek out higher-performing schools that can deliver a better education for their children. It's one of the most common arguments for expanding school choice. But is it true? Is that what really happens? What’s the connection between school quality and interest among families in those schools?
Kindergarten enrollment patterns in New Orleans give us a pretty good test case. All families must choose an elementary school – there are no default, zoned assignments. Nearly all schools participate in a universal enrollment process called OneApp that collects their preferences and issues assignments using an algorithm. OneApp releases detailed reports showing how often each school is selected as a family’s first choice as well as how often each school is listed somewhere on a family’s rank order.
Do families prefer higher performing schools?
The most widely cited measure of school quality in New Orleans is a letter grade given by the state. As we’ve discussed before, letter grades are almost entirely based on student proficiency rates on state exams, plus a small component for helping non-proficient students advance to proficiency. Each school gets a School Performance Score (SPS) that maps to a letter grade.
Schools with higher SPS results attract higher demand.* Take Ben Franklin Elementary – also known as “Baby Ben.” It’s a B-rated school, with the sixth highest 2014-15 SPS of any K-8 school participating in OneApp. For the 2015-16 school year, it was listed on 412 kindergarten applications for the main assignment round. All of its seats were filled immediately. The school with the lowest SPS, William J Fischer, is an F-rated school. It was listed just 42 times. It opened the school year with unfilled kindergarten seats.
Below, you can see a scatterplot of SPS scores versus frequency of appearance on kindergarten applications.
So there we have it, right? Quality and demand are functioning in relatively strong harmony. Parents are identifying the better schools and flocking to them.
Not so fast. While SPS and demand show a relationship, it’s worth asking whether SPS is the best measure of school quality. We’ve written before about our preference for school ratings that consider student growth, not just test pass rates. In our view, a good school is one that moves students forward over time – no matter where they started.
When we use EdNavigator’s revised school performance scores, we get a slightly different picture. The relationship between school quality and demand is now weaker, with a higher number of outlier schools that don’t fit the pattern as neatly.**
For example, consider Harriet Tubman and Medard Nelson. Tubman is one of the highest growth schools in New Orleans. In EdNavigator’s adjusted school grades, Tubman earned a B – and very nearly an A—and it was named on 59 kindergarten applications in 2015. Medard Nelson, on the other hand, is one of the lowest performing schools in New Orleans, with an F grade from the state and EdNavigator alike. And yet it was named on 61 applications – more than Tubman.
This is not an isolated instance. Schools with very similar performance levels show dramatic variation in popularity. And schools with very similar popularity show dramatic variation in performance.
Baby Ben, the most popular school in town, performs comparably to Mary McLeod Bethune, but Bethune was named less than half as often. Martin Behrman, a relatively low growth school that earned a D from EdNavigator, had the second-most applications on the Westbank, trailing only ultra-popular Alice Harte. Scan the graph for yourself and you’ll see loads of other instances.
What does it all mean?
First, as we’ve discussed before, there’s more than one definition of school quality, and parents aren’t always using the same definition as the state – or the same definition as EdNavigator. Decisions that appear irrational on paper may be perfectly rational for a family weighing many different factors, from how far the school is from home to what sorts of sports it offers. Academic performance is hardly the only thing they’re considering. We can’t forget that.
Second, we’ve learned that just pumping out information is not enough to fuel a healthy system of school choice. Families need help. They need access to impartial, trusted advice that helps them balance their personal priorities and constraints with the options available to them. They need to know how the assignment system works. They need time to learn about schools they don’t know much about.
At EdNavigator, we wholeheartedly support school choice. It helps ensure that students are never locked into a school that is struggling or a poor fit simply because their families can’t afford to move. But everyone interested in giving families more school options should be careful not to presume that the equation for success is information plus opportunity. If that’s all we’re offering, we shouldn’t be surprised when things don’t add up.
* The correlation is about 0.56, for the stats junkies out there.
** Again for the stats folks, the correlation is about 0.27.