Does Going To a Good High School Help Get You Into College?
In San Francisco, the answer appears to be no.
Introduction
It’s college application time. And high school application time. And the future of Lowell’s admission process is being debated again. Three good reasons to have a look at the classic question of whether your choice of high school affects your chances of getting in to a good college.
I’m not concerned here with the quality of the education at high schools, just with whether the name of the high school gives a boost at college application times.
Since most private college application data is jealously guarded and is more heavily influenced by non-academic factors such as parental wealth, legacy status, and athletic talent, I’m going to focus on the University of California and in particular on its two most selective campuses: Berkeley and UCLA.
How Many Get In
The vast majority of the best students will apply to one or both, either as a first choice or as a backup. Each year, these campuses both receives application from between 1600 and 1800 San Francisco high school students. UCLA typically accepts around 200 and Berkeley around 225. 2020 was a particularly good year with over 340 acceptances at Berkeley and 235 at UCLA. Some students are admitted to both so it’s not public how many distinct students are admitted.
Here’s the number of admissions by type of high school. The numbers fluctuate a lot from year to year so I’ve taken a five-year average.
Private schools, whether independent or religious, have only about one-quarter of San Francisco’s students but they account for over 40% of successful applications. Lowell by itself is nearly another 25%.
If you really want to see the breakdown by school (and you know you do), here it is:
Glancing at these figures (Lick-Wilmerding with 120 seniors gets more students admitted than Lincoln with 500) and concluding that the school matters is appealing but overly hasty: we haven’t yet adjusted for applicant quality.
If a school is favored by the admissions office, its students will require a lower GPA for acceptance or will get accepted at a higher rate than equally qualified students from other schools. Is this true?
Applications and Admissions Per Student
Lowell is by far the largest high school in the city - of course it has the most students admitted. Let’s look at the results on a per student basis. The University of California only publishes admissions statistics if at least three students from a high school were admitted. Only ten San Francisco schools met this threshold for both campuses in all five years. I’ve added some other schools that met the threshold in most years but since they have the fewest reported admits, the data for those schools (i.e. Bay, Lycée Francais, Balboa, Urban) should be treated with particular skepticism. Other schools had to be excluded entirely for lack of data.
This shows how many students submitted applications to Berkeley and UCLA and how many were accepted as a percentage of the number of 12th graders in each school. To smooth over the year-to-year fluctuations, I’ve taken the five year averages in all cases.
At the prestigious schools, more students apply to the top UCs and more students get in. Lowell students submitted a total of 4033 applications to either Berkeley or UCLA in this period and received 552 admissions (i.e. 13.7% of applications were successful). Lowell had an average of 693 12th graders each year so that’s 4033 / (5 * 693) = 1.16 applications per student and 552 / (5 * 693) = 0.16 admissions per student. Many (most?) students who applied to one would have applied to the other so the 1.16 applications per student means that at least 58% of Lowell students applied to these schools and somewhere between 8% and 16% (depending on the number who were admitted to both) of Lowell students were admitted to one or both. In comparison, Saint Ignatius students submitted fewer applications (0.88 per student) but the number of acceptances per student was almost identical (0.156 per student).
How Good Are The Applicants?
Submitting an application merely shows ambition. It doesn’t measure the quality of the application. UC also publishes data on the average GPAs of those who apply, those who are admitted, and those who ultimately enroll. We can compare the average GPA of those who apply with the success rate of their applications.
As we would expect, the higher the applicants’ average GPA the greater their chance of being admitted. It’s clear that there’s an almost linear relationship and that most schools lie close to the trend line. If a school lies away from the trend line, that may indicate that there are too few admits for the data to be reliable (e.g. Bay, Lycée Francais, Balboa) or it may be that they are particularly favored by admissions offices.
Mission is particularly notable: most San Francisco high schools got slightly more students admitted to Berkeley than UCLA but at Mission the ratio was over 3:1. In both 2018 and 2019, Mission had more than twice as many students admitted to Berkeley as Lincoln, Galileo, and Washington even though it is one third smaller than each of them.
How Good Are The Admits?
Although the average GPAs of the applicants ranged from below 3.80 to above 4.00, the average GPAs of the students who are actually admitted to either Berkeley or UCLA cluster remarkably close to 4.20, regardless of high school. A span of 0.10 GPA points covers all the San Francisco high schools bar Mission and Bay. In particular, students from the more prestigious high schools seem to have no advantage for admissions. The prestigious high schools get more students admitted because their students are better qualified on average not because admissions officers favor those schools.
UC has a very particular way to calculate high school GPA: it considers only 10th and 11th grades and caps at two per year the number of honors-level courses for which a student gets credit1. This means the maximum GPA an applicant can have is 4.33. (It also means that a student who takes two AP classes and four regular classes and gets an A in all will have a higher GPA for UC purposes than a student who takes six AP classes and gets an A in five of them but a B in the sixth. Which is weird)
About 76% of all CA public school admits to UCLA and 69% of public school admits to Berkeley had GPAs over 4.20. But fewer than 40% of applicants with a GPA over 4.2 were admitted to UCLA and fewer than 45% to Berkeley. In comparison, students with a GPA of 3.80 to 4.19 had only a 7% chance of admission to UCLA and a 13% chance of admission to Berkeley.
A better predictor of applicant success might not be average GPA but proportion of students above 4.20. I’d expect Lowell to have a higher proportion of such students than the other schools. Given that Lowell’s acceptance rate is exactly in line with what the average applicant GPA would predict, this suggests that Lowell students might even be at a disadvantage.
AP Exams As a Predictor
We may not know the proportion of applicants with GPAs over 4.20 but we do know how many AP exams were taken at each public school and what scores were obtained on those exams. AP exams are scored on a 1-5 scale with a score of 3 or higher considered a passing grade.
For all the AP exams sat in SFUSD schools in the period 2016-2019 (I excluded 2020 because the pandemic disrupted them that year), the following chart shows:
the percentage of all SFUSD AP exams that were taken at the school
the percentage of all SFUSD AP exam passes were obtained at the school
the percentage of all scores of 5 (the highest mark) were obtained at the school
and compares these to:
the percentage of admissions to Berkeley and UCLA achieved by students of the school.
Lowell students sat 42% of all the AP exams taken by SFUSD students; they obtained 53% of all the AP exam passes and 66% of all scores of 5. But they achieved only 44% of all admissions to Berkeley and UCLA.
The next largest schools, Washington, Lincoln, and Galileo, also had fewer students admitted than their AP participation rates would suggest. Their students took 32% of the AP exams, obtained 29% of the AP exam passes and 22% of the scores of 5 and achieved 23% of the Berkeley/UCLA admissions.
SOTA students accounted for 7% of all AP exams taken, 6% of all AP exam passes, and 4% of all scores of 5 but achieved 9% of all the admissions to Berkeley and UCLA. Those students presumably score very highly on the extracurricular part of the application evaluation because of their Art prowess.
Burton, Wallenberg, and Mission together account for 11% of AP exams taken, 7% of AP exam passes, and just 4% of all scores of 5. Nevertheless, students from those schools obtained 19% of all the admissions to Berkeley and UCLA.
Without access to detailed student records, we can only speculate about why this might be:
UC gives extra weight to students who are the first in their families to attend college or who tell stories in their applications of overcoming adversity. There are probably more of those students at the east-side high schools.
UC evaluates applicants in the context of their high school and the opportunities available there. A good student might find it easier to shine at those schools than at, say, Lowell. Perhaps it’s easier to get an A. Perhaps excellent recommendations (“the best student in my class”) are easier to obtain.
It’s not a race thing, at least not the way you might be thinking.
At Wallenberg, 29 out of 34 students accepted to Berkeley and 10 out of 14 accepted to UCLA were Asian (ignoring the years when too few students were admitted for any breakdown by race to be published).
At Burton, 23 out of 30 students accepted to Berkeley and 12 out of 14 accepted to UCLA were Asian (again, ignoring those years when no breakdown by race was published)
At Mission, 28 out of 71 admitted to Berkeley were Hispanic/Latino and 32 out of 70 were Asian (the denominators are different because I only include the denominator for years when I know the numerator and the numerator isn’t published if it’s lower than 3). 16 of 28 admitted to UCLA were Hispanic/Latino.
The Decline of Feeder Schools
It not be true today that attending a prestigious high school increases your chances of getting in to a top UC. But it was true in the past. Schools acquire their reputations for a reason. The earliest five-year span for which UC publishes admissions data in 1994-1998. Here is the data for that period compared with the most recent data.
In 1994-1998, San Francisco high schoolers received nearly 2900 acceptances to Berkeley and UCLA and students from just 4 schools (Lowell, Saint Ignatius, University, and Lick-Wilmerding) received 69% of them. In the most recent five-year period, total admissions from San Francisco are down 21% to under 2300 (UC has become far more competitive and San Francisco’s population hasn’t grown) but the number admitted from the Big 4 is down by 47% and they now constitute only 46% of San Francisco admissions to Berkeley and UCLA.
Conclusions
Don’t choose a prestigious high school because you think it’ll give your child an advantage when applying to the top UCs. It won’t.
Choose the high school that you think will be the best match for your child. The prestigious high schools remain excellent schools. If one of them is the best match for your child, go for it. But don’t stress if your child doesn’t get in. Students whose grades are good enough can get in to the top UCs, regardless of their high school.
If you do want to maximize your child’s chances of getting in to a top UC, consider one of the east-side public schools.
More precisely, a student can earn no more than 8 semester credits for honors classes. That’s equivalent to two year-long AP or Honors classes in 10th grade and two more in 11th grade.