Wednesday, December 3, 2014

How to Take Socioeconomic Status into Account in Redistricting?

At last night's Policy and Engagement Committee meeting, the ICCSD administration informed the committee and the community that they had received a cease and desist order concerning the Diversity Policy from the State Board of Education based on feedback from the USDA. See the Press-Citizen report (here). In the USDA letter, it becomes clear that it is illegal to use free and reduced lunch status to, for example, create density maps for the purpose of redistricting. A necessary part of redistricting.

As I have argued previously, the particular provisions in the policy are (a) not being implemented currently and (b) would have required substantial revision in order to be workable. So, I'm not sure what losing the Diversity Policy really means at this point. Still, I believe it would be best for the Board of Education to think more about the end goal they want to create (reducing the achievement gap, reducing barriers to education in target populations, improving the academic success of all students particularly those who are least advantaged, or some other such goal) and to let the administration work out the details of how to realize that goal. I argued for that previously (here). Not only do I think that would work better, but it also exemplified how the board and administration should interact. See this Press-Citizen opinion piece about that (here) and this blog post on the topic (here).

But, like many others, I also strongly believe that socioeconomic status must be taken into account in order to deal with the challenges we face as a district (see here for an argument about that). Other approaches considered independently of socioeconomics (e.g., using English-Language Learner status, race, and/or standardized test scores) have significant problems that I won't go into here. I'm happy to talk about that in comments or later, if need be.

But this made me wonder how to address socioeconomic status without using free-and-reduced-lunch status? Here's my tentative answer:

First, take the average household income data that is readily available. Unsurprisingly, there is a strongly positive correlation between household income and FRL status. This is a graphic from the 2013 Dejong-Richter report on demographics in the ICCSD. We may need a slightly more fine-grained analysis of average household income (that is, smaller blocks and also taking into account great wealth, too).

Second, take the student density maps (here). If the average household income was as fine-grained as this density map, then you could easily draw pretty good redistricting lines that would take into account socioeconomic status. 

Third, I don't think it is wise to specify exactly how socioeconomic status should be equalized through redistricting. As I mentioned above, I think we should characterize our goal in terms of an educational outcome, and then give the administration latitude to realize that goal. 

But how might such a method work: first, you take the density map. Let's say that the block with 234 students on the density map is going to Grant Wood. Then we could multiply 234 by $50K. And let's say that the block  with 37 students near Shimek is districted to that school. Then, we multiply 37 by (at least) $100K. Now, we can see what we'd need to do with redistricting in order to bring the average household income closer together at those schools. In fact, this method could very well be a significant improvement over using FRL status, since it would take into account the uber-wealthy and some low-income persons who choose not to apply for FRL.

Of course, I'm assuming that the average income information would basically be true for the population of families with school age children. I think that assumption probably holds well enough presently for us to use something akin to this approach as a tool in the overall approach that the administration would take. I think the assumption may not apply in all circumstances, and that's another good reason for the board to focus on goals and allow the administration to use this tool among others to accomplish those goals.

Now, don't hear what I'm not saying. I don't think this should be the only means for redistricting. Nor do I think that it is a cure all for what ails our district. Rather, socioeconomic considerations need to be in the administration's tool belt for addressing the barriers that low-income persons disproportionately face, which is exacerbated at school with high concentrations of poverty.

What say you? Is this approach sufficient to take socioeconomic status into account? Or do we just have to use FRL status? 


  1. The idea of mandating change VS. incenting change is an issue/problem. I have a problem with both FRL and your approach because they are both based on the idea of forcing people to change schools based on parental income.

    This focus on parental income is not beneficial and is delaying progress in other areas. How much time and energy has been wasted, how much ill will created, because of our current DP? We MUST create win-win situations. Magnet schools and voluntary movement can be part of the solution. Forced bussing based on family income is not the way to go.

    People like neighborhood schools – we must find solutions to the achievement gap involving choice and changes like improved facilities and smaller classes -ones that people actually want – and that they would be willing to change schools for.

    1. Thanks for your comment. Still, I don't think I specified an approach, so I'm not sure what you mean by "my approach." I specified a goal, and suggested that the means used to accomplish that goal should be suited to the task. I didn't say what the specific implementation should be (magnet, incentivizing movement, sister-schools, redistricting, increased resources, some combination of these and other options, or whatever).

      I do believe that socioeconomic considerations would need to play some role in accomplishing the goal, but the particular implementation is based on whether it helps us reach the end goal better than alternatives. If magnets are the best way to do it, then let us do that. If redistricting some based on income is part of the solution, then let us do that. My guess is that the administration would need to use different tools in different contexts to accomplish the goals. But I think the board should get out of the business of specifying particular modes of implementation ASAP.

  2. So, I honestly don't see the point of redistricting based on anything except population density and school size. Granted, I don't know much about geographical politics and education law, so I may be way off base. But I had always assumed that city services like roads, power lines, water mains, fire departments, hospitals, and schools were intended to all provide reliable and effective service to the surrounding geographical community, and to be distributed in such a way that the whole city is served.

    If a particular power substation isn't working properly, the solution isn't to redraw the utility infrastructure such that the affected population can still get power. Yes, that might work temporarily, but it's a hack. The long-term solution is to fix the malfunctioning substation. Likewise, I don't see how gerrymandering the school district map can help address the root problem of schools which aren't properly serving their local communities. Might the school board's cease and desist simply be a stern reminder to focus on improving schools rather than shuffling children around?

    1. Josiah: Thanks for your comment. I think a good number of folks think about educational systems like you suggest.

      For me, I think the educational structures should be situated such that it (a) provides at least a basic minimum of education to all (I leave it open what that is), (b) that, independent of a, differences in the structure should benefit those most in need rather than those who already possess the greatest resources, but (c) we should be mindful and hesitant when efforts to make things better require significant structural changes that may, in fact, cause things to get worse.

      In the status quo, I would say that the basic minimum is provided at all of our schools sufficiently well. Furthermore, I think it is clear that our current educational structures add to, rather than ameliorate, the structural injustices that make it significantly harder for those who are least well off. I've described how our system makes it worse off in the third part of this post.

      So, what does that mean for your response? It means that I think your analogy is inapt. First, no one is defending "gerrymandering" (well, except the status quo, which is pretty well gerrymandered in such a way as to benefit those who are rather affluent--cf. the North Lincoln island and the Windsor Ridge neighborhood). Second, I don't think the problem is that we need to "improve schools." We attract some of the best teachers in the state, and overall, our schools are excellent. Third, your analogy assumes a single provider of education in a community like a power company. That's not the case.

      I think a more apt electrical analogy would be a circuit breaker system. If you have a single breaker than keeps getting flipped (or a fuse box where the fuse continually gets blown). (a) You can change the wiring to allow greater electric current (that would analogous to providing more resources in target schools?).(b) You can plug fewer devices into the relevant sockets moving them to other ones (this would be like redistricting?). (c) You can rework the internal wiring between the plugs and the various breakers (maybe this is a combination approach of redistricting and increased resources as appropriate?).

      I don't see any of those things as a "hack". Rather, they are responses to a specific problem that arises in that specific context.

    2. Thanks Michael! I always appreciate your thoughtful posts and patient responses to folks like me who don't quite have the full picture. :-)

  3. I agree that the diversity policy's goals were too specific, especially given that there was no discussion in advance of what it might take to meet those goals. But while it might be fine to let the administration come up with the details, I do think that any plan that would involve boundary changes (or, for example, the creation of magnet schools) should have to get board approval before being implemented.

    With that qualification, I'd agree that the board should have given a broader charge to the administration and then waited until there was a concrete proposal on the table before debating it further. As it turned out, an awful lot of energy went into debating the details of this Diversity Policy, only for us all to find out two years later (a) that they can't use FRL numbers, and (b) that meeting the targets they set by the required deadlines might create more problems than it would solve.

    1. I agree it should go before the board. But I'd also say that if the position is reasonably aimed at the intended goal, then the board should be deferential to the administration on such matters while holding them accountable if they fail to actually accomplish the goal as directed.

  4. The latest cease and desist order for Diversity Policy shouldn't shock anyone considering the Iowa Dept of Agriculture and USDA told the school district that the policy would violate federal law back in March 2014. Murley and the board chose to ignore those warnings and continue on their path which unfortunately wasted time and resources. I agree completely with Josiah's comments. Schools should be redistricted based on population density and growth only with students attending schools closest to their homes and provide additional resources to those schools that need them. Balancing socioeconomic status is theoretically a very noble goal but would be very difficult if not impossible to implement as a long term solution for many reasons.

    1. Anon 7:16, thanks for your comment. However, I disagree with the claim that population, growth, location, and school size should be the only factors in school assignment. I think those are important, and their importance shouldn't be minimized. That is, I disagree with those who say that "walkability" isn't valuable; or that "walkability" is achieved if you are within two miles of one's school assignment. I also think we need to build to address capacity concerns. But there are other considerations as well, and my guess is the best solution for improving the academic achievement of the least advantaged populations in our district is a mix of principled redistricting and principled resource allocation.

      For what it is worth, I think there is very good reason to think that this is the dominant view of our community.

      Furthermore, it would be impossible (given school size constraints) to have all students attend the school closest to them. That would mean you'd probably have about 500-600 at Kirkwood, around 700 or 800 at Wickham, around 700-800 at Wood, and it would likely be even more problematic at the Junior High and High School levels. Now, I'm not saying that distance from the school shouldn't be a factor that is considered in redistricting. It should be, but my guess is that any prescriptive policy would have significant problems.

      As a humorous aside, one concerned parents once proposed that the school district adopt a policy that required students be assigned to a school within one-mile of their residence. I joked that such a policy would entail that I am required to send my kids to 5 different schools. A friend suggested that I only needed to have one more kid to be able to send one kid to each elementary within a mile of where we live!

  5. Josiah - the problem is that the power substation analogy doesn't really capture the dynamic in this situation. A substation operates in the same way, with the same efficiency, no matter what geographic area that we put it down in. That's not the case for a school. Teaching and learning are both work very differently in a school with a high poverty concentration than they do in a more affluent school. Students who come from impoverished backgrounds, who don't get regular meals or regular sleep, whose parents are too busy working low wage jobs to be involved in the school or to provide meaningful education-related help at home, face barriers that middle class and affluent students don't. Concentrating large numbers of those students in a single school, as we do in some of our schools where upwards of 70% of the students qualify for free and reduced lunch, concentrates those problems and makes those schools work less efficiently, especially for those kids who need the most help. This isn't a case of a "malfunctioning" school, as any school dropped into that geographic zone is going to function the exact same way.

    Its also not a static problem. Once a school reaches a tipping point, which is usually somewhere in the high 50% FRL range, it tends to accelerate, as parents of means transfer out, or move out of the neighborhood. And, its exacerbated by the disparities in what kind of housing is available in particular neighborhoods. If there aren't attractive housing options available for middle class or affluent families in a particular school's attendance zone, then those families will seek housing elsewhere and attend a different school, and the high poverty concentration at the school will affect the market value of the housing that is in the school, further incentivizing those who can to move elsewhere. All of this works together to put schools in zones that become more and more economically isolated. If you want to stick with the power plant substation analogy you have to imagine substations with vastly different capabilities, and whole school zones where there are regular brownouts and blackouts and service interruptions, not because of any malfunction, but because of basic structural differences that result from the geography that they sit in. And you have to imagine that those substations become progressively less powerful as time goes on. Grant Wood school's population was at around 40% FRL in 2008. Its now at 79%.

  6. And, contra Anonymous above, there are a couple of reasons why it’s very hard to solve this problem with "more resources." One is that it takes a tremendous amount of resources, specifically in the form of classroom teachers, to make a difference in a school with a high concentration of poverty, and those resources are at a premium in this district right now. We have fairly high class sizes right now, and just instituted awful budget cuts across the district. If you put the teachers required into the schools that need it, then class sizes are going to grow to an alarming degree in the more affluent schools. I'm skeptical that this will even be pursued with the necessary vigor, given that parents in the more affluent schools have the time and energy and social capital to become squeaky wheels in ways that this district has generally responded to, and in ways that most parents in the high poverty schools don't. In any case, redistricting to achieve demographic balance is actually a lot less expensive than putting enough resources into these schools to actually make a difference.

    The other problem with trying to solve this problem with additional resources is that its, as Josiah termed redistricting, a hack. It doesn't solve the underlying problem that creates the disparities and the inefficiencies. It doesn't do much to break the cycle that leads to some zones becoming increasingly economically isolated. And, because of this, that commitment of resources has to be ongoing, delivered year after year. This is not only expensive, it puts the district in the position of institutionalizing these disparities.

    So, looking more thoroughly at this, balancing the socioeconomic demographics of the schools isn't just a noble goal, its a less expensive and more sustainable way to fulfill a public school system's basic commitment to deliver the same educational opportunities to every kid, regardless of their wealth or race or address.

  7. In addition to using data in a way that was not allowed, the current DP had major issues; it is good that it is dead.

    Now we can concentrate on coming up with something that is better.

    While there will be more, here are a few ideas for the next go around:

    • How about something designed by people who know what the hell they are doing?
    • How about something with academic goals vs. capacity requirements?
    • How about a plan that focuses on the elementary level because that is where the largest issues are (and fixing those helps take care of the higher levels)?
    • How about something that uses magnets and incentives for movement vs. forced bussing?
    • How about something that listens to the parents and teachers from the schools we are actually trying to help?

    1. I think these are all reasonable points (if a bit harsh at times), and related to points I've made frequently.

    2. I'm definitely down for a better and more workable solution, and those are all reasonable points to me, barring a couple of caveats:

      1. "forced busing" is a meaninglessly loaded term. You're assigned to a school, which to some degree is "forced" I guess. But you're no more or less forced to go to a new school than you are to go to the one you're currently assigned. If you live outside of two miles, you qualify for a bus. You're not however forced to ride that bus if you can get to school otherwise. This is true whether you're going to the school you are currently assigned to, or if you find yourself assigned to a new school. so, if there isn't a more specific definition to be had, then all of the 42% of the kids in this district who ride busses are the victims of "forced bussing."

      I understand of course that the point here is to make redistricting seem onerous by invoking a lack of choice and by invoking "busing" which has negative connotations associated with attempts to desegregate larger urban environments. But its a pretty sloppy and loaded way to get at a question that needs to be asked: to what degree would a given desegregation/redistricting plan cause an uptick in busing and to what are the consequences (expense to the district, travel time to students) of such an increase? The proposed maps this year included some increase in busing, though we never learned exactly how much for any of them. I think that some increase in bussing would be less expensive than the level of resource commitment, but student distance from school was definitely part of the problem in some of the maps we saw. I think that in most cases the concerns about busing were overstated and imprecise and leveraged unfairly as indictments of the whole idea of redistricting for balance.

    3. I do think that its good that the District is pursuing Magnet options, but its also worth keeping in mind that putting a successful magnet into place takes a good, long amount of lead time. And that's time that we continue to allow a significant number of kids to be educationally underserved.

  8. I'd also add that I'm a parent at one of those schools that this is designed to help, and that a number of our parents and teachers are in favor of seeking a solution via redistricting and have been very supportive of the Diversity Policy, while still understanding that it was flawed, and still finding issues with the maps that were put together to try to implement it.