Some Charter Operators Outperform LAUSD, Others Do Not

Green Dot. Alliance. Aspire. PUC. Magnolia Science Academy. Celerity. KIPP.

These are major charter operators in Los Angeles – each operating 6 or more schools.

But not only are they large operators – they are outspoken ones. They are the ones asking the district for co-locations, they are the ones who petitioned for Public School Choice. They are the ones who are the subject of documentaries about the charter school revolution.

But, unfortunately, because of the way that charter schools are regulated, it is difficult to look at them like a unified district.

When the state reports their test scores, each of the charter operator’s schools are reported as individual districts. Aspire Centennial is its own district. Aspire Titan is its own district. Aspire Pacific is its own district. So you cannot see how these charter operators perform as a whole charter network.

Until today.

To make the charter schools into districts, I am going to try a slightly different methodology.

Up until now, when I have looked at the data for schools, I have used medians to measure to performance across different schools. The problem is – the median ignores the size of the school. A school with 160 students has the same weight as school with 1,000 students.

In order to avoid this error, I have used the CAASPP research file to calculate approximately* how many students there are total who are in each category from an entire charter network. It is sort of like combining all of the charter network’s students into one big school.

I should also note that I am only looking at L.A. county schools. Aspire has schools in NorCal, Green Dot has schools in New York and Tenessee, PUC has a school in Rochester (Yes, Rochester), and KIPP has schools everywhere. I am not counting those schools. Only L.A. County schools.

First, let’s look at the English scores:

Source: CAASPP Research File

KIPP and Celerity impressively lead the pack here. LAUSD non-charter schools are the lowest performing overall in English.

However, I find it interesting how small the difference is between LAUSD and most charter networks. Aspire, Green Dot, PUC, Magnolia Science and Camino all have scores within 8% of LAUSD – not really a significant gain.

This is only reinforced when you look at Math data.

Source: CAASPP Research File

LAUSD actually outperforms Green Dot, Alliance and Magnolia Science are in Math. Green Dot is a full 10% lower than LAUSD’s non-charter schools. Frequent readers of my blog may question whether I am trolling Green Dot. I have had several posts that have not been flattering for them. But this is just what the data says. Over and over and over again.

PUC, Camino and Aspire barely beat the district. That does not sit well. LAUSD has empowered private operators to run schools in Los Angeles with the hope that they can do things better. What is the point of having charter operators who make only marginal improvements?

Back in the data, yet again, Celerity and KIPP are high achievers.

KIPP and Celerity are different than the other schools in a couple of ways. First, they only run elementary and middle schools. They also usually (not always) use a Open-Upwards model – opening one grade at a time so they mostly have kids who have gone through their entire system. That gives them the opportunity to really hammer in their school culture and deepen their model.

Is that what it takes to make schools more successful in Los Angeles? Starting from scratch and building upwards?

I think that starting from scratch definitely helps when you are trying to replicate a model. And replication is one of the key factors that makes charter networks attractive. If they can’t replicate their success, then they negate their own purpose.  And if they don’t outperform the district – then what is that purpose anyway?


Note: LAUSD non-charter data includes a lot of schools that are wealthier and from more affluent parts of Los Angeles. But it also includes Special Education Centers and Continuation Schools. I predict that the factors balance out – but I will be sure to look more into that data on a future date.

 

 

 

4 Replies to “Some Charter Operators Outperform LAUSD, Others Do Not”

  1. I’m wondering about grade levels. Shouldn’t you compare all the charters that are elementary schools to LAUSD elementary schools, etc?

  2. Thanks for the question. Many charters operate non-traditional grade spans. For example, many KIPPs operate grades 5-8. Does that make it middle or elementary? Or a K-8 – should it count as two schools? I am staying out of that question.
    L0oking grade by grade, I have found, is also VERY convoluted and hard to understand. Unless you account for multiple factors, it can end up being a measure of their original starting level, instead of their growth. I am trying to create a regression for growth, but its harder than it seems. Stay tuned.

  3. Freddie deBoer says:

    Isn’t the biggest issue the fact that we have tons of reasons to believe that students are non-randomly distributed between different types of schools?

    1. School Data Nerd says:

      We definitely have a lot of reasons to believe this. However, when it comes to available demographic data, most charter schools closely resemble their local district schools. I wrote a post on that, you can search for it. But there are few exceptions. 1. We have terrible data on parent education. For some reason, a significant portion of LAUSD parents opt to report “Decline to state” for education, while that percentage is very low in charter schools. That may be because they have lower education, or because they have a survey problem. 2. There are ton of non-quantifiable variables when it comes to student information. For example – parents who have their students in charter schools (or magnet schools) took the step to apply to a school. Perhaps that shows a level of involvement that is not matched in local schools (although speaking from experience in charter schools, it generally doesn’t). More importantly, parent involvement is generally not quantifiable in public data.

Comments are closed.