A surfeit of reality

The gingerness with which the researchers approach their unwelcome findings makes it a little difficult to determine their outcomes.  They do get to it, but they wrap it all in wordy gauze.  From Increasing Access to Selective High Schools through Place-Based Affirmative Action: Unintended Consequences by Lisa Barrow, Lauren Sartain, and Marisa de la Torre.  From the Abstract:

We investigate whether elite Chicago public high schools differentially benefit high-achieving students from more and less affluent neighborhoods. Chicago’s place-based affirmative action policy allocates seats based on achievement and neighborhood socioeconomic status (SES). Using regression discontinuity design (RDD), we find that these schools do not raise test scores overall, but students are generally more positive about their high school experiences. For students from low-SES neighborhoods, we estimate negative effects on grades and the probability of attending a selective college. We present suggestive evidence that these findings for students from low-SES neighborhoods are driven by the negative effect of relative achievement ranking. 

Chicago has Selective Enrollment Hight Schools (SEHS) which admit based on the most gifted citywide.  However they reserve slots for students in their local catchment area who score the highest but do not attain the minimum.  The questions being explored are: Do SEHS improve student test scores? Are they able to raise low SES affirmative action admit scores by a greater degree (implying that quality of school, quality of peers, and quality of teaching have some ameliorative impact above simple IQ)?  Do affirmative action admits improve their chances of attending selective universities?  These are all the outcomes desired and used to justify affirmative action admits.  

But the answers are No, No, and No.

IQ is a superior forecaster of education attainment than SES or quality of school attended.  Some of the findings buried in the verbiage.

The affirmative action admissions policy in Chicago, reserving seats for students from low-SES neighborhoods, makes selective schools more racially diverse than many other high schools in the city.

We find no evidence that SEHSs are raising test scores more for students from the most disadvantaged neighborhoods than for students from the most advantaged neighborhoods.

SEHSs have a positive effect on students’ perceptions of the high school experience.

SEHS students are more positive than their counterparts in non-SEHSs.

SEHS students are more likely to say that students get along well and treat each other with respect.

Students in SEHSs also report a greater sense of personal safety; namely, they are less likely to worry about crime, violence, and bullying at school. 

We find negative effects of being admitted to an SEHS on GPA. This effect is primarily driven by the large negative impact on GPA for students from more disadvantaged neighborhoods. 

Students admitted from the lowest-SES neighborhoods are, on average, the lowest-achieving students in selective schools. 

Admission to an SEHS reduces the probability that a student from a low-SES neighborhood attends a selective college.

One of the most critical findings is buried in the text.  

Students from low-SES neighborhoods who are admitted to a selective high school are less likely to attend a selective college than students from low-SES neighborhoods who just miss the admissions cutoff. 

A less PC rendering of their findings might be:

 Chicago has a high school affirmative action program.

It achieves more racial diversity in the school system as a whole and students feel good about their SEHS experience.

SEHS teaching does not improve education outcomes above that predicted by student IQ results.

Affirmative Action admits score lower grades because they are less capable than their SEHS peers, putting them at the bottom of the grade distribution curve.

If you are a lower SES facing the choice of affirmative action or staying at your local school, your chances of attending a selective university are greater by attending your local school than going with SEHS based on affirmative Action.   

Even more bluntly, affirmative action is serving a diversity goal and a feel good goal but does so by harming lower SES education attainment.

Even though the researchers try and hedge their position, this is pretty strong evidence of the academic mismatch hypothesis.  

There is an awful lot of reality in here which the education establishment really don't want to see or here.