Home > musing > The modeling death spiral for public schools

The modeling death spiral for public schools

May 16, 2012

There was recently a New York Times article about how the public schools have become super segregated by race.

I’m wondering how much of this can be explained by income rather than by race in combination with the obsession we all have with test scores. Let me explain.

If I’m living in a neighborhood with a neighborhood school and the school seems pretty good, then depending on how picky I am I might just stay living there and let my kids go there.

Now assume that suddenly there are test scores available for all the schools in the area, and it turns out my neighborhood school doesn’t do as well as a surrounding neighborhood. Then, depending on how much I think those test scores matter to my childrens’ futures, and how much resources I have, I will be tempted to move to that neighborhood for the “better schools” (read: better test scores).

Over time, people with good resources will move to the new neighborhood, which will become more expensive because there’s competition to get it, which in turn will make it easier for that town to raise local taxes to improve the school, and will also attract parents who really care about the quality of the schools, which will improve the school and presumably the test scores of that school, exacerbating the original difference of test scores.

And of course that’s just what’s happened in this country. My parents moved to Lexington Massachusetts for the schools, and they paid a premium for their house for the location and the school system. So I went to a public school but one that increasingly was attended by richer and richer kids.

Income segregated public schools are the new private schools.

In New York City, where there is more to consider than just your neighborhood, because you can get your kids into schools in other neighborhoods, and there’s a whole network of gifted and talented schools as well, it’s a much more complicated dynamic, but the underlying reasons are the same, and they again have to do with segmentation modeling: we know which schools do well on tests and we avoid poorly testing schools if we can.

The availability of the test scores is huge- if I’m thinking of moving to a new city I can just look up the SAT scores of the high schools in the area and try to find a place to live which is in one of the highest-scoring towns.

This is what I call a death spiral of modeling, and it’s the same idea I described here when insurance companies have too much information about you and deny you coverage because you need insurance so bad. And it’s very difficult to get out of a death spiral, because to do so you need to reset the whole system and re-pool resources but in this case people have already moved out of town.

Questions I am thinking about:

  • Is it dumb to care so much about test scores? On the one hand I don’t want to take chances on my kids, so I will opt for the conservative route, which is to think they should be surrounded by kids who test well, because certainly in extreme cases that kind of thing is likely to be contagious behavior. But maybe we have exaggerated ideas about how contagious these things are or how important test scores really are to our kids futures. How would we test that and how would we disseminate the results? And what if we found out that everybody has been acting totally rationally?
  • Which begs the other question, namely how can we get this system to work better overall for the average student that would be realistic?
  • Note that in the above discussion I haven’t talked about the teachers at all, which is strange. But from my perspective, our system is all about concentrating kids who test well together, and it’s not all that clear that the teachers matter, although I’m sure they do actually. What am I missing? Is there a way of solving this death spiral problem through awesome teachers?
Categories: musing
  1. ARC
    May 16, 2012 at 7:27 am

    This is very well-written and exactly what I see happening in the area around me (an affluent suburb). Even within a single school district there is a flight to the top-testing schools. but each school has the same per-pupil budget, and same criteria for hiring/evaluating teachers e.g. teachers are hired by the district, not by a specific school – so its not even a resource issue *within* a school district. The test score differences essentially come from more educated (and typically richer) parents putting their kids in the higher testing schools. The quality of education isn’t explicitly different, but arguably your kids benefit from being around smarter peers. The other question is whether teacher or peers matter more to what your kids learn. As for solving, I have no idea. Putting my kids in a top public school is my compromise solution for wanting them to go a public school, yet still ensure they receive a high-quality education.


  2. May 16, 2012 at 8:38 am

    I certainly agree with the basic problem that folks move to better school districts, impoverishing schools that are not doing well, and it’s hard to solve that problem through teachers or any other means. But have test scores really exacerbated this? I remember when I was a kid my parents knew where the good schools were, and they were recent immigrants with few connections in the area beyond my father’s work.


  3. Brian
    May 16, 2012 at 9:00 am

    This is only true assuming perfect information and zero barriers to changing schools.


    • Jay
      May 18, 2012 at 1:27 pm

      Actually, if there were zero barriers to changing schools, then poor people would be able to go to high-performing schools as easily as rich people, and we wouldn’t have this problem. Instead, the problem is that the system rewards people with the resources to overcome the barriers, making the rich better off at the expense of the poor.

      Test scores online don’t count as perfect information, but they’re available and actionable, so good enough.

      It’s amazing how often the attempt to measure performance screws it up. There are always ways to score cheap points on the measurement while leaving other aspects of the job undone. The measurement transforms from a way to measure how well the job is done into the reason that the job must be done poorly.


  4. beterday
    May 16, 2012 at 10:01 am

    When you add in the teachers the problem becomes only worse. My daughter teaches in a very low income school in Austin, TX. The kids are all poor and non-white and all of the teachers, excepting her and one or two others, are looking for new jobs in less challenging situations. The work of teaching is easier in a rich school as you do not have the behavior problems so the best teachers flow to the best schools; the pay is the same or better and the best schools will recruit the best teachers. That leaves the poor school with the ones not hired by the rich schools. Now you are not only working with the most challenging children, but the least motivated and creative teachers and the death spiral accelerates. Pay incentives have to be given to work in the lower performing schools and in fairness, my daughter will have some or her federal student loan debt paid off if she continues to work in the poor school. If the government were more involved, not less, in the financing of college they would be better able to incentivise working at inner city schools.


  5. JSE
    May 16, 2012 at 10:57 am

    “Pay incentives have to be given to work in the lower performing schools”

    And five years later you can count on hearing your state legislature say “lower-paid teachers are correlated with higher test scores, so we should pay ALL teachers less in order to fix the public schools!”


  6. May 16, 2012 at 11:13 am

    Most parents want to find a school environment where their kids can thrive. That’s not always the place with the best test scores. It might be a school with a great softball reputation. More than anything, parents want a good community for their kid. Friends that encourage good behaviors and build self-confidence with social skills. If the school is not up to par in an important area, it’s usually possible to supplement sports, activities, and academics in other ways.

    Even at highly diverse schools, the social networks among students tend to break down racially and then economically within racial groups. Interracial friendships are a bit more common at these schools but still relatively rare – at least from what I’ve seen in North Carolina. Over the past generation, there appears to have been fewer social interactions between white-collar and blue-collar kids. This probably reflects changes in the neighborhoods themselves and the schools simply reflect the increasing segregation of the working classes.

    It’s not clear to me that an average kid would necessarily be better off at the academic bottom of the class at a school with high scores instead of being at the top of the class at a school with low scores. Expectations need to be carefully managed in both situations.


  7. Dan L
    May 16, 2012 at 11:45 am

    Income segregation and “test score segregation” are real phenomena, but I think you are downplaying the role played by good old fashion race segregation. I’ll focus on NYC, since it’s the school system I understand best. Many whites would never dream of sending their child to a school where whites are the minority, no matter the test scores. As evidence for my claim, I present TAG, which is a school for gifted children in east Harlem that is only 3% white. Although it is undoubtedly a high-performing school and is reasonable commuting distance from UES, it is shunned by white families. While I know there are legitimate criticisms of that school, race is surely a big factor. For exhibit B, I present two elementary schools in Chinatown that also run gifted programs. These two are consistently considered to be the least desirable gifted programs in their district (at least among whites), and I can’t imagine that being 90% Asian has nothing to do with why.

    Also, I totally disagree when people say or imply that moving to a better a school district/zone is “like going to private school.” No, it’s not. It’s nothing like it.


    • JSE
      May 16, 2012 at 11:50 am

      “Also, I totally disagree when people say or imply that moving to a better a school district/zone is “like going to private school.” No, it’s not. It’s nothing like it.”

      I dunno, Cathy and I both went to excellent public schools in affluent neighborhoods and both of us think it’s something like going to private school. Not identical — but like.


    • Cathy H
      May 17, 2012 at 11:04 pm

      I agree race segregation is still a big part of the problem. Looking at income and test score segregation we can not help but notice racial injustice. Furthermore, there are the examples Dan cites where white people still don’t want their children to attend majority non-whites schools, despite their good performance. However, paying attention to the issue of income and test score segregation is a good place to start and does get a the practical root of the issue if not the moral and psychological root.


  8. Solving Education
    May 16, 2012 at 12:53 pm

    Technology could solve this problem by allowing the BEST Teachers (teachers from schools with the highest standardized test scores) to teach classes via the internet and make these classes available to ANYONE ! Suddenly, students trapped in failing-schools and failing school systems would have access to COMPETITIVE Courses that tend to produce high-standardized test scores. To personalize Online Learning to each students needs, Computer Adaptive Technology can be used to pinpoint their strengths and weaknesses. Unfortunately, Education is HIGHLY POLITICIZED and has more to do with maintaining INEQUITIES than creating equal opportunities. This explains why inner-city students are practically forced to receive inferior educations from failing schools. For example, Tonya McDowell got 5-years in prison for attempting to send her son to a higher performing school outside of her “district” (she was homeless at the time in question). Her accusers say Ms. McDowell “stole a free education” that is only “free” to residents of the school district. The value of the “stolen” educational benefits is about 15K or 1-year in prison for every 3K in educational benefits. Applying this ratio to Criminal Wall Street Bankers and Corrupt Politicians who collaborated to steal our Wealth and Prosperity through Fraudulent Mortgages, Securities Fraud, Market Manipulation, Counterfeiting etc, they should receive prison sentences of about 10-Billion years (using a 30-trillion dollar estimate of stolen wealth and prosperity) http://www.dailymail.co.uk/news/article-2108733/Homeless-mother-Tanya-McDowell-sent-son-6-better-school-wrong-town-jailed-years.html


  9. May 16, 2012 at 1:18 pm

    You just discovered (kind of) Schelling’s chessboard model of segregation:



  10. Ray
    May 16, 2012 at 2:11 pm

    A big part of the underlying problem is that public schools are funded by local property taxes. This is just obscenely regressive and seems manifestly unjust to me. The constitutionality of this is being litigated in New Hampshire in a series of cases that have been dragging on for 20 or so years now, but there hasn’t been any actual change. There just doesn’t seem to be the political will to fix this, which I find somewhat astonishing and very depressing.

    But I’m fairly certain that even if we were able to live the pinko dream of funding public education federally with a progressive income tax, and even if it worked perfectly and eliminated any funding disparities that we would view as unfair or unjust…the geographical distribution of educational outcomes (defined as standardized test scores or really whatever metric you want) would still be highly skewed.

    School funding is just one of the multitude of variables that determine educational outcomes. A big part – maybe the biggest part – of the variation in student outcomes can be explained by what’s going on at home. Every teacher I’ve ever asked about this agrees about the crucial importance of these social-capital-type factors.

    I suspect that families that care enough (and are wealthy enough) to uproot and move to a better school district generally would do a fine job of raising and educating their kids in any district. (I’d definitely watch a reality TV series that tested this suspicion by plucking some public-school-emigre family out of, say, Lexington, and put them in New York and enrolled their kid at Murry Bergtraum or someplace and documented their lives.)

    Regardless of where they’re living or what’s going on in the classroom, these kinds of families are probably going to be paying closer attention to their kids’ homework, or buying and making their kids work through supplemental enrichment materials, or enrolling their kids in summer math camps. And furthermore, just by being themselves, they’re also going to be doing all sorts of passive things that aren’t necessarily designed to educate their kids, but nevertheless do socialize their kids to value education and academic achievement – leisure reading, weekend museum trips, dinner table conversations about heady scholarly topics.

    I would expect that these families are also going to tend to cluster geographically for reasons totally unrelated to public school performance. The same traits that correlate with high child educational achievement probably also correlate strongly with income. So even if you remove the feedback-loop effect created by the fact that rich people want to live near schools that produce good outcomes, they’re still going to end up living close to each other. And consequently, the schools that happen to be near the places where they live are going to produce good outcomes. (This reminds of trying to prove that all Cauchy sequences are convergent, en route to constructing the real numbers in Math 161.)

    I guess my point is that the income segregation effect is actually independent of the availability of information about outcomes. Not sure why I had to ramble on for so long to make that point – slow day at work.

    So it seems to me like the income segregation problem is actually a consequence of basic human social tendencies that we as a liberal market society would find really hard to control. But that doesn’t mean that we shouldn’t at least try to build some sense of decency into the things that are easier to control, like the distribution of public funds. Even if equalizing funding won’t completely solve the underlying problem, I strongly suspect that there are a lot of kids on the margins for whom it will still improve things a lot. Maybe we can’t cure the cancer, but we should at least try to quit smoking.


  11. May 16, 2012 at 4:23 pm

    I did terribly on my SATs. Like under 1000 bad. Now its all MCAS study, which isn’t good imo, because what does that prepare you for other than taking the MCAS?


  12. HungryGhost
    May 16, 2012 at 6:53 pm

    With a lot of controversy, Texas has dealt with this issue by making state college entrance based almost entirely on high school rank, and thus an indirect form of affirmative action. It’s called the Top 10% plan, although for UT Austin the legislature had to relent and make it more like a Top 8% plan, in order to leave some room for private schools, international students, and out of state applicants.

    The side effects is that you’ve kids at UT Austin from poor high schools (including rural whites) who are at their top of their class but only have 900/1600 SAT scores sitting next to kids with 1500/1600 kids from high powered elite suburban high schools. Kids just below the Top 8% at great suburban schools increasingly have to go out of state, even when they have fantastic SATs and APs and many merits. I know one person who graduated from Westlake High School but was accepted at UC Berkeley but rejected by UT Austin.

    In addition, Texas has an explicit affirmative action system, but that is currently being litigated at the US Supreme Court.

    The Top 10% system will likely be permanent because it’s a strange bedfellow arrangement of rural and urban state senators being more powerful than suburban state representatives.

    Interestingly, liberal California has no system of either type.


    • May 16, 2012 at 7:32 pm

      I’d be interested to know what happens to the kids with mediocre test scores at UT Austin. Do they flunk out or drop out at a much higher rate than others? Can they access programs with competitive admissions or are they herded into programs designed for athletes? Are they ostracized by their peers? My guess is that most are poorly served and might be better off at another school.


      • HungryGhost
        May 16, 2012 at 8:14 pm

        They do have lower GPAs but there are enough bullshit majors that the university flunk out rate isn’t that high overall. Flunk outs are usually people who party and don’t go to class, and Top 10% in any high school are those who at least get their work done. (Flunk out in engineering/comp sci is of course high, but people just switch to communications studies or something of similar banality.)


  1. No trackbacks yet.
Comments are closed.
%d bloggers like this: