Tuesday, September 2, 2014

Should we stay (with more resources), or should we go (on a bus)?

The Iowa City Community School District has been having more meetings about redistricting, the diversity policy, magnet schools, and attendance area. I've been doing some preliminary evaluation of data from the 2012-2013 school year as a way to give community members insight into the relative costs of various proposals that are currently on the table.

I have previously expressed my support for the spirit of the diversity policy (here), and I am strongly committed to that same spirit. I am not committed to the specific percentages, guidelines, time-frame, or any particular strategy for accomplishing the spirit of the diversity policy. I'm open to various possibilities, but I am not open to continuing the status quo, nor do I simply want change for change's sake. I want meaningful change that is intended to address the structural problems harming our educational system for our least-advantaged students. 

A few months ago, I offered a survey (here) mainly as a way to help our community think about the various possibilities related to redistricting. I identified to dominant strategies that could have some effect on reducing burdens that our community has directly and indirectly places on low-income students. 

Strategy One (from the survey): "Allocating additional resources to a school over a long period of time (especially more teachers to have a better student to teacher ratio) is the best way to remove educational barriers to learning that disproportionately affect low-income students, which results in the district having no need, or at the least, very little need for boundary changes."

A number of respondents expressed sympathy for this approach. I wanted to crunch the numbers to get a better indication of the actual costs associated with using this method as the primary vehicle for removing those educational barriers.

Relevant assumptions and limitations:
  • You would need class sizes at or below 15 students at the Kindergarten level (17-19 is recommended for the average Kindergarten classroom), 19 students for 1st and 2nd grade, and 23 students from 3rd-6th grade. (It may be that class sizes would need to be lower than this to truly make a difference, but I needed to make an assumption that seemed reasonable.)
  • I only considered increased resources at 3 schools: Kirkwood, Twain, and Wood. We have some others elementary schools that are well above 50% for their FRL rate, and they may have to be included if current trends continue.
  • Class sizes for those three schools and others for 2012-2013 can be found here (see page 165 in the pdf file).
  • Based on the official class sizes for 2012-2013 at those three schools and the suggested class sizes above, you would need 4 more teachers at Kirkwood and Twain, and 3 more teachers at Wood.
  • Any changes to increase resources at those three schools would be budget neutral, i.e., there would be tradeoffs with other programs or funds.
  • The tradeoffs would come from the 11 elementary schools that have FRL rates lower than 30%.
  • The cost of reassigning a teacher (salary and benefits) would average at least $75K. Teacher salaries for 2014-2015 can be found here (page 158 in the pdf file), and the information there suggests that the assumption is a fair one. I am, however, open to other figures if they are more compelling.
Given these assumptions, the strategy one model (i.e., the increased resources model) would need to reassign the 11 teachers (or an equivalent amount of resources to affect a similar outcome) from the 11 elementary schools with FRL rates lower than 30%. Thoughts, questions, or concerns about just using Strategy One?

Strategy Two (from the survey): "Boundary changes that achieve socioeconomic integration for all our schools is the best way to remove educational barriers to learning that disproportionately affect low-income student, and there would be little or no need to allocate additional resources to particular schools since they would be balanced socioeconomically."

A number of respondents expressed sympathy for this approach. So, how does the number crunching work for this model?

Relevant assumptions and limitations:

  • There would need to be extensive changes to the attendance areas at Twain, Wood, and Kirkwood in order to use this method. 
  • I used the number of buses in 2012-2013, which are here and here. The total number of buses at all levels of 102.
  • The per bus cost is difficult to calculate. We are charged a lump sum from a 3rd party for our busing services, divided between the general education fund and other funds. Approximately 2/3 dollars spent on busing is not from the general education fund, leaving only about 1/3 of the cost coming form the general education fund.
  • Our primary concern is with general education costs, as it has the most significant impact on the classroom, and it was the reason behind the budget cuts last Spring. See here for my discussion of that.
  • The total general education costs for busing in 2012-2013 was a little less than $4 million.
  • That figure divided by 102 results in a cost of just a little less than $40K per bus.
  • The number of students transported in buses varies considerably, with some carrying as many as 75 students as evidenced here.
  • The number of buses at Kirkwood, Twain, and Wood vary (see here). Twain had 4 in 2012-2013, Wood had 1, and Kirkwood had 0. 
  • I only considered increased resources at 3 schools: Kirkwood, Twain, and Wood. We have some others elementary schools that are well above 50% for their FRL rate, and they may have to be included if current trends continue.
  • Any changes to increase resources at those three schools would be budget neutral, i.e., there would be tradeoffs with other programs or funds.
  • The tradeoffs would come from the 11 elementary schools that have FRL rates lower than 30%.
 Given these assumptions and guidelines, I estimated that Strategy Two (redistricting only) we would need between 3-5 new buses. At least 2-3 at Kirkwood, at least 1-2 additional buses at Wood, and probably no additional ones for Twain (given that they already have 4). At a cost of $40K per bus, that would be the cost equivalent of about 0.2 teachers taken from each of the 11 elementary schools with FRL rates under 30%. Thoughts, questions, or concerns about just using Strategy Two?

My Commentary:

I appreciate that people offering both of these strategies are dissatisfied with the present state of affairs. Our system is unjust and disproportionately burdens low-income students. That needs to be addressed, and I think our community has the political will do to so now.

Still, I think there are significant challenges that we have in using either of these methods as the only overall solution. There is no realistic way of getting Wood and Twain under 50% FRL given the constraints that the administration placed on the process and approved by the board (the cluster constraints, mainly). Furthermore, given that so many students, with a high-concentration of poverty, live around Kirkwood, it poses its own challenges for both models. 

Can you fit more teachers there without redistricting? Are other resources able to improve academic achievement as well as lower class sizes? How much are we harming low-income students if we redistrict them to a school that is significantly further away from their closest school?

In light of these challenges and more, I think it would be best to pursue a hybrid approach of these two strategies -- leaning more to the redistricting side of the equation unless there are significant overriding reasons to not do so. To not do so, would have bad results (i.e., increased class sizes) for a majority of our school district's students. Despite what common sense may suggest, it is significantly more costly to pursue Strategy One exclusively rather than Strategy Two exclusively.

Furthermore, the communal context may dictate that you take different strategies with different schools. Twain's already heavy bus load may make it such that redistricting is more attractive there. Kirkwood's capacity limits may mean that we will have to redistrict some of its students to other schools, while increasing its resources? Perhaps Alexander will spur real estate development, which would help reduce the ultimate need for more resources at Wood and Alexander? Perhaps the administration needs more latitude to accomplish the goal of addressing injustice in our system, while we hold them accountable for the results.

Thoughts, questions, or concerns about my commentary?


  1. Michael, Thanks for doing this. I mean, really, Thank You. Why the administration isn't doing this sort of analysis, I will never understand. One piece of the puzzle that I am wondering about... if we reassign 11 teachers to those three schools, can you estimate what impact that will have on class sizes at the low FRL schools? How big will classes get there in order to get smaller sizes at the high FRL schools?

  2. Anon 10:17: It is hard to know, and it would vary depending on the size of the school. Certainly, it would mean that you have to relax the aspirational goals for those schools for a grade. My guess is that average class sizes would probably not be affected so much (because it would likely only be one grade out of 7) but the percentage of classes that are out of compliance with the aspirational goals would go up substantially. I'd have to crunch those numbers to get those results though.

  3. As a district, we should learn from the successes and failures other areas have already had in attacking the achievement gap. Our current DP does not have educational goals, but rather it has capacity and FRL percentage goals – this is not a recipe for success.

    I would strongly encourage you to push for a hybrid approach that minimizes forced busing and encourages choice as much as possible (which I believe is also what the grandfather of the FRL integration approach also strongly recommends).

    1. I agree with you for the most part. Although we may have a different idea of what "forced busing" means.

    2. Michael,

      Forced busing is forcing a child to change schools and busing that child to a farther school against a family's wishes due to FRL status VS. capacity. Meetings all around the district have shown parents at both ends of the FRL spectrum are against this. Making adjustments that are reasonable as new schools are opened is a better option.

      The achievement gap is the true problem we really need to address, yet capacity differences and high levels of poverty in a school are the problems our DP seeks to address. There is a difference; there may be a relationship, but poverty levels are only a part of the problem. I think our current board is on the right track.

      Researching what the most successful districts are already doing to attack the achievement gap and attempting to emulate it makes the most sense.

    3. Anon 8:22: So, is it forced busing if a student lives 2.3 miles from their closest school, and that student is redistricted for FRL reasons to a school that is 2.4 miles from their location but the parents would prefer to stay at their current school? What if they were redistricted for FRL reasons and because a new school opened? I'm not sure what your answer to these questions would be, and that suggests the definition is poor. So, I still don't understand what you really mean by forced busing.

      And lest you think this is a fanciful example, there are students in our district bused over 3 miles to a school, and at least one proposal suggested moving them to a school that was slightly further way but not by much for FRL reasons (and a new building was coming online as well). The parents were opposed to the move. Is that forced busing?

      Furthermore, historically the language of "forced busing" has referred to taking inner city urban kids and sending them to the outer suburbs of a large urban area while sending those outer suburbs into the inner city. I don't know anyone who is advocating that sort of proposal here. No one is saying that we should send kids around Wood to Van Allen, for instance.

      I think this sort of careful discussion is important, as I think having vague values will push us into absurdities if we aren't careful about how we use them.

      But here is where I think you are right: we should reduce busing as feasible and, on balance, it is preferable to go to closer schools rather than schools that are further away. But you have to weigh the good versus the bad in any policy matter, and there are times when policy makers should choose to send kids to schools that are close but still further way than their current school, and sometimes injustice in our educational system is a sufficient reason to do that. But there are times when it increases injustice to do that, too. So, we have to weigh the good and the bad.

      In any case, I hope that clarifies my point. That's what I was driving at in my initial comment.

  4. I was hoping the Board would not just relax the constraints of the diversity policy but also the artificial constraints of the current cluster designations. For instance, in Cluster 1, it would make a LOT more sense for Borlaug to be a receiving school of FRL kids from Kirkwood (especially the island identified in the islands option currently on the table), and then some of Borlaug's FRL kids from married student housing could then to go to Lincoln. The university's new married student housing complex is REALLY close to Lincoln, an easy walk for those parents if they needed to walk to get there. Likewise, it would be an easier walk for Kirkwood parents in the current island options to get to Borlaug than to Wickham. There are so may ways to tinker with these maps that aren't being attempted for some reason.

    Likewise in Cluster 2, there are options there that haven't been investigated. I really think the paired school option makes more sense there because there are so many small schools so close to each other. This would also result in more equal class sizes across all the grade levels across the entire district and would save $$$ by allowing for class sizes bigger than are currently averaged on the east side of the district in the upper grades in particular.

    I am for a better socio-economic balanced school district, but within reason. Island bussing is not reasonable. Fixing the instances where kids are already being bussed past one school (or more) to go to another school is entirely reasonable. So North Ridge kids currently being bussed past Kirkwood to go to Central should no longer be going to Central. Likewise in Cluster 2 there is a similar instance of a neighborhood being bussed past other schools to go to Longfellow I believe.

    Another part of being reasonable is NOT redistricting an area until a new school/major addition opens up which necessitates redistricting at that time. For North Liberty schools especially & Wickham, redistricting in 2015 and then likely redistricting again for the new North Elementary school opening which I cannot fathom occurring later than 2017 would place an undue burden on those kids. That elementary school needs to be built before 2019 - as I think it is impossible to ignore the growth data after this fall's enrollment data is considered.

    We can keep our neighborhood schools, and even improve upon that concept, while better balancing FRL across the district.

    1. Lots of good thoughts here. I have my own thoughts about how redistricting should be done, and they align somewhat with what you are saying here. My main rule of thumb with regard to that is to start with the end in mind, and start making strides in that direction as soon as you can. I see a lot of that in what you say, and that's a good thing.