At long last, the final report on the first year of implementation of the new student assignment system is out. The Ad Hoc Committee on Student Assignment will discuss the report on Monday evening, but for the wonks among you, here it is ahead of Monday’s meeting — 80+ pages of maps, charts and data to dig into and analyze. Leave me your thoughts in the comments!
Some light weekend reading . . .
This entry was posted in student assignment and tagged data, recap, student assignment. Bookmark the permalink.
Rachel Norton 
Something else I wanted to add: It seems that this assignment process had similar results to the old one (in terms of racial/ethnic and SES diversity). But it’s also a way simpler system for parents to navigate. The tactics of “trying for 0/7” and all that are now gone.
However, it’s clear from the chatter that transparency is still an issue. Clarity around siblings is still bad, and it now looks like how the “swap” process works is not well-understood by the districts customers (the parents). It’d be great to have some statistics around swaps: how many were there district-wide and per-school. Stuff like how many people swapped out of, say, Grattan. And how many swapped into Feinstein. It’d be good to see that the number of swaps district-wide is stable from one year to the next.
It does seem that a good tactic would be to put high-demand schools lower on your list to increase the chance of getting swapped into a higher choice (e.g. put Rooftop down as #8 in the hopes of getting it and being swapped into De Avila).
“Bernal Dad: I disagree with your interpretation, and the report’s, that making neighborhood school allocation a higher priority wouldn’t have made much of a difference (I voted no on Prop. H, BTW). I find figure 10.39 revealing: green indicates families who was assigned to a school outside their AA that was ranked lower on their list than their AA school. Simply as a factual statement, there is a lot of green.”
Looking at Figure 10.39 and Figure 18, I only see 4% that were enrolled in a choice school that was ranked below their closest school. 4% isn’t “a lot”, IMHO. You may be focusing in on a limited number of schools (Stevenson, Sutro, Jefferson, Clarendon) where there’s significant numbers who were enrolled in a choice school that was ranked below their closest school. (Note closest school and Attendance Area school aren’t necessarily the same in all cases.)
In general, I’ll also note that a choice system needs a lot of slack capacity if everyone is to get one of their choices. Last year, the applications were 107% of capacity. So there’s no conceivable way for everyone to get one of their choices, or apparently even to inflect the numbers who get one of the choices up significantly. Also, as the numbers who list their AA school *at all* is only 48%, and only 25% listed their AA school as #1, a strict neighborhood assignment instead of a choice system per the Prop H’ers desires would make the situation worse than the current assignment system.
We’re done here. Except for the middle school feeder system, the information in this report makes me believe there’s no further optimization of the Student Assignment System that’s going to yield more people getting one of their choices.
However, underlying the angst that will be on display because of the Assignment Letters, we should note there’s a success story there – that SF schools are more and more operating at capacity.
As a Bernal parent who toured Serra, I have to say I’m similarly disappointed that Serra isn’t “ready” to get to the next level. It’s on my list, because for a school I don’t *love* I’d rather not travel across town, but my strategy if we end up there is to try to make it work, but expect to enter the lottery again next year. Sad but true. The parents who have stuck it out report that “with just a few more neighborhood families ready to roll up their sleeves and put in a lot of time, things could change” and “it’s a lot better than it used to be.” Item #1 won’t work for me — with a full time job, I’m able to do *some* behind the scenes work, but can’t be *in* the school much at all, and Item #2 does not instill confidence — not when I can drive 15-20 minutes to get to some schools I LOVED that will be a good fit for my child.
@Bernal Dad: I disagree with your interpretation, and the report’s, that making neighborhood school allocation a higher priority wouldn’t have made much of a difference (I voted no on Prop. H, BTW). I find figure 10.39 revealing: green indicates families who was assigned to a school outside their AA that was ranked lower on their list than their AA school. Simply as a factual statement, there is a lot of green.
I don’t know how much green would still be there if sibling preference remained higher than AA, but I think 10.39 goes to your statement much more than 10.36.
@Rachel: I appreciate your running this blog. One broad comment as a public educator new to San Francisco: I feel that the discussions and reports focus almost entirely on “redistribution”, i.e., who goes to Clarendon and who goes to Muir. This is an important question, but there is very little discussion of the costs to the system of the loss of many kids who on average probably have motivated and politically influential parents.
For example, 15% “other white” enrollment in middle school seems like a dramatic failure that doesn’t receive nearly enough attention in reports like this. I refuse to believe that the parents who move to the suburbs or go with private school would do so independent of their public school options, but there seems to be a lot of blaming of those parents on this and other blogs.
The response is usually that it could be worse: we aren’t DC or Baltimore. I think we should have higher goals. If nothing else will motivate the bureaucrats of SFUSD, you might point out that it will be increasingly difficult to justify high levels of school funding if more and more politically active parents are not attracted to public schools.
As a specific question, is there any chance of a study by the group of academics who designed the system of how different assignment choices affect the total student population? Since the Board dismissed them, they probably have time on their hands.
Also, is there any prospect of their being invited back into the evaluation process?
“For me, the most shocking stats were the racial breakdowns of applicants vs. district wide enrollment.
K12 enrollment in October 2011 (p. 17) is 12% caucasian
K applicant pool for 2011-2012 year is 26% caucasian”
The numbers of enrollees in K/6/9 showed a bit less defection of caucasians out of the system, though, than in the past: 15% caucasian enrollees. So that’s a bit more diverse than the current intake.
@Frank: “The fact that 25% of the families who don’t enroll actually got their first choice is an eye-opening statistic.”
About 55% of those who don’t enroll got one of their choices. I think it indicates that for at least 25%, and maybe as high as 55% of those who don’t enroll, publics were their Plan B for if they didn’t get into the private school of their choice.
“On Junipero Serra, I’ve met one of the families who tried going there. The expectation from the existing families are just so different from the “takeover” families, that there was no easy way to change the school to make them comfortable. Many, many families didn’t go to first grade.”
That’s very disappointing. I’d hoped that Serra would go a similar path to what seems to be happening with Glen Park and Sunnyside with being embraced by all the community there. Evidently it’s not going to happen anytime soon.
@Rachel On siblings, Rachel Norton got the breakdowns on this latst year or two years ago. The difference was very small (a couple of percentage points). There’s clearly a perception among parents that this is a bigger deal than it is. That said, the District could go a long way to improve the perception of this if it called out the sibling effect in these reports — even if the effect is very small.
@Bernal Dad On Junipero Serra, I’ve met one of the families who tried going there. The expectation from the existing families are just so different from the “takeover” families, that there was no easy way to change the school to make them comfortable. Many, many families didn’t go to first grade.
The fact that 25% of the families who don’t enroll actually got their first choice is an eye-opening statistic.
The ads are put there by WordPress, I have no control over them and receive no revenue.
Why are you running ads on your website?
Is school readiness ever measured or accounted for by SFUSD in looking at school assignment and enrollment? SFUSD has many low-cost or free preschool programs and most middle class parents send their children to preschool before kindergarten. I’m wondering what percentage of children enter kindergarten with preschool experience and how that factors into enrollment statistics.
For me, the most shocking stats were the racial breakdowns of applicants vs. district wide enrollment.
K12 enrollment in October 2011 (p. 17) is 12% caucasian
K applicant pool for 2011-2012 year is 26% caucasian
Wow.
Anyway, this is a very juicy document. It’s interesting that despite the change to the assignment system, there really wasn’t a whole lot of change, nor would there be if AA preference was prioritized over CTIP1.
Other interesting facts: 25% of those who don’t enroll got their first choice, and a majority of those who didn’t enroll got one of their choices. So much for the assignment system chasing middle-class parents away – most of those who go private or suburban got one of their choices, a quarter their first choice, but still declined.
Also: Junipero Serra’s racial stats a total surprise. I thought there was a movement amongst white middle-class parents in Bernal to support Serra, but zero white kinder enrollment there in 2011, down from 12% in 2010, is a surprise.
“It’s sad — I just realized that at my day job we run more reports like this in a few minutes than you guys do in a year.”
Likely using canned data into canned CRM software into canned reports. Come on, this ain’t comparable.
Wow. Figure 10.36 is a real refutation to the neighborhood schools boosters.
I really wish these reports would separate out the number of sibling applicants from non-sibling applicants to give more clarity to the choice statistic (ie: “63% of kindergartners received their first choice” — but what percent of those were siblings? same for 2nd, 3rd choices, and so on).
@ Wayne – I would recommend reading the report as the data isn’t organized in exactly the way you’ve asked the question, but I think the best answers would be found in the section starting on page 32.
Rachel, I have the same question that I’ve been asking for a couple of years, first as a prediction and later as an analysis of the data:
For a Clarendon attendance area non-sibling student, what are the odds of getting assigned to Clarendon for GE assuming that it was the first choice?
Also :
For a Clarendon attendance area non-sibling student, what are the odds of getting assigned to Clarendon or a higher choice assuming it was placed as a choice?
Thanks
Wayne
thanks for posting Rachel. How interesting to see the statistics of how many actually apply to the assignment area schools or even their closest schools (which can be different.).
Page 55 finally tells the story — no choice at all for some neighborhoods (surprisingly Cobb is one of the worst due to lack of CTIP1, which masked most of the unpopular schools).
It’s sad — I just realized that at my day job we run more reports like this in a few minutes than you guys do in a year.
This is such a fascinating document. Thank you for posting it! I am curious whether the board is planning to discuss the possibility of setting a cap on the percentage of CTIP students enrolled in each school that is not in a CTIP area. This would enable more neighborhood students to get into their schools, and it could increase diversity at other schools. For example, at some of the schools in the center of the city, the breakdown could be evened out between the schools. Clarendon, Rooftop, Alvarado, and Sanchez all have over 30% CTIP students; whereas Grattan, McKinley, and Milk have between 13% and 18%. My son attends Grattan, where the diversity could be improved. On the other hand, a friend of his in the Clarendon district couldn’t get a spot there because such a small percentage of spots went to neighborhood kids. If the cap were set at 20% or 25% then some of the CTIP students who got spots at Clarendon (which was presumably their first choice) could be assigned to Grattan (which might be their second or third choice). This would even things out a bit for everyone. Do you know if something like this is being considered?
An amazing amount of data. Well very organized and easy to get through, in spite of the size. I’m a little surprised at the % of people who did not list their neighborhood school as the first choice. I live in the Miraloma area and I have yet to meet a parent who didn’t list it first. According to the data, less than 50% listed it first. Odd.
Also, I would have liked to have seen some sort of analysis of Citywide school data. Immersion seems to be growing more and more popular. Unless I missed it, there seems to be no mention of Citywide schools whatsover.
I’ve just skimmed through it, but it looks like there is nothing at all about assignment for students who require special education services.
Why am I not surprised? It is always an afterthought, down at EPC.