My second entry in my Benchmark Project comparing the Grosse Pointe Public School System with Ann Arbor, Birmingham, Bloomfield, Farmington, Northville, Novi, Plymouth-Canton, Rochester, and Troy.
Innumerable variables affect Michigan K-12 finances, but we need to start somewhere. We’ll attempt to organize as much data as we can (while still breathing) to assess the health and trajectory of Grosse Pointe and comparable districts and surrounding communities.
I selected ten Metro Detroit suburban districts reasonably similar in some ways. We might think of these communities in similar markets (i.e. where a family may live if based in the Metro Detroit area). Doing so makes analysis more manageable and hopefully more meaningful. I have included the following benchmark districts:
- Ann Arbor Public Schools
- Birmingham Public Schools
- Bloomfield Hills Public Schools
- Farmington Public Schools
- Grosse Pointe Public Schools
- Northville Public Schools
- Novi Community Schools
- Plymouth-Canton Community Schools
- Rochester Community Schools
- Troy School District
To start, I combined Michigan Department of Education data with U.S. Census (or rather American Community Survey) data. We’ll look specifically at population (using school district boundaries) and demographic data, specifically the school aged population (ages 5 through 17) and proportional measures to explore some key topics:
- Since school funding follows enrollment and enrollment is largely (but not exclusively!) affected by population, we’ll cross reference enrollment, population, and demographics. Below you’ll see what proportion of a community population is school aged and how does that compare with actual enrollment. These are reasonable proxies for market opportunity and marketshare.
- Population trends will have the greatest impact on enrollment, but it can be a sticky thicket. A community’s population may increase, but the proportion of school aged children may decrease. We’ll pay close attention to these breakdowns.
- In Michigan we have the complexity of School Choice, which allows a school district to enroll non-resident students. If a school system experiences enrollment limitations due to political/district boundaries, it can increase its market opportunity through Schools of Choice.
The resulting analyses will expose problems and opportunities. Over time we’ll pay particular attention to community demographic data and school data. I am hopeful my entries here will increase our collective appetite for quantifiable inputs more than anecdotal – a particular hazard of our social media era.
Enough preamble. Let’s wade in.
We start (slide 1) with Student Enrollment by benchmark district. Most districts enjoyed an increase in enrollment from 1992 through the mid and late oughts (00s) at which time Michigan’s Great Recession made landfall, inciting reduced growth rate or enrollment contractions – a painful fate. By and large, however, this benchmark group in total has held reasonably close to the peak levels just preceding The Great Recession.
(Click on the images to expand the view.)
There’s life before and after the Great Recession. In slide 2 we use 2005 as a demarcation point, isolating Student Enrollment trends from 1992 to 2005 and then from 2005 to 2019. The earlier period, which closely coincides with the start of Michigan’s Proposal A era, was generally kind to most districts. Growth in all respects was common, making life much easier. In fact each of the Benchmark Districts grew in that time (some tremendously).
Things got real after 2005, after which none of the ten benchmark districts grew more than the previous period and six experienced enrollment decline – some far outpacing others, namely Farmington (-32%), Grosse Pointe (-18%), and Bloomfield Hills (-10%).
Slide 3 introduces other community demographic data, namely total population and percentage of population that is school aged. We start with the averages of the ten Benchmark Districts and then each individually. I suggest comparing individual district (slides 4 – 13) with the group averages on slide 3.
The Great Recession halted total state population growth, but the total population of the Benchmark Group increased. But on the downside, the school aged proportion of the total population dropped slightly. Enrollment as a percent of School Age Population, however, increased dramatically. This could be partially attributable to public school enrollment taking share from non-public schools (as household budgets tightened), but a better guess is that School of Choice enrollment drew share from both benchmark and non-benchmark group districts. We can see examples of this in the breakout slides below.
Ann Arbor Schools (slide 4) enjoyed population gains as it grew in economic and demographic consequence with the digital economy. But Ann Arbor’s school age population proportion has declined over the last 15 years. However, the Enrollment as a percentage of School Age population shows that Ann Arbor did one (or both) of two things: Increased their share of enrollment among its population and/or attracted enrollment from other districts. In future posts we can explore this. Also of note in Ann Arbor we see that their School Age percentage is significantly lower than the Benchmark Group average revealing that Ann Arbor residents (not surprisingly) are less likely to have school aged children. These are some of the data nuances to examine for each benchmark district.
In Birmingham (slide 5) we see a population that contracted from 1995 to 2005, but has recovered slightly since then. They’ve held reasonably steady on their school aged population proportion and have increased their marketshare slowly but steadily from 1995 to now.
In Bloomfield Hills’ case (slide 6) we see a slightly contracting population made worse by a decline in proportion of school aged children. But with a 30 year high in their marketshare (87%), they are likely holding on to their school age eligible population and have likely enhanced that position by their attractive limited schools of choice option – the International (IB) Academy.
Farmington (slide 7) has held steady in population, but have lost mightily due to a drop in both school aged proportion and, more menacingly, a serious drop in marketshare from 2005 (which was a healthy 87%) to 76% in 2019.
In Grosse Pointe enrollment has declined steadily from 2005 (down 18%), outpaced only by Farmington (-32%) among Benchmark Districts in that time period. Many smaller pieces(rather than one big one contribute to GPPSS’ struggles – population declines, lower proportion of school aged children, and marketshare loss. We’ll spend more time with GPPSS data later in this series, but it may be that demographics (combined with no school of choice) demands a closer look. It’s a key question uniquely for GPPSS. With 86% marketshare in 2019 compared to 87% in 1995 (all free of school of choice noise), one could argue that the GPPSS is capturing their normal share, but overall community demographics (aging and declining population) are hard to overcome. The question then becomes how to respond to this dynamic.
In Northville meanwhile (slide 9), we see a community growing steadily in population, holding steady in proportion of school aged population, but particularly benefitting from school of choice, which would explain how their Enrollment of school aged population has been at or above 100% throughout the Proposal A era. EDIT: I have done further research after this initial guess about School of Choice Enrollment and I was dead wrong. I will be posting about School of Choice numbers shortly, but Northville has extremely limited dependence on school choice. In that same update, I’ll theorize on why Enrollment % of School Age Population could be above 100%.
Novi (slide 10) is similar to Northville – steady population growth, generally steady school aged population, and above 100% on Enrollment percentage of school aged population – again probably attributable to schools of choice. Novi may be a beneficiary of Farmington’s loss of share. EDIT: See note above on Northville. Novi also has limited dependence on open enrollment.
One of the larger suburban districts, Plymouth-Canton (slide 11) has also experienced steady population gains with slight losses in percentage of school age children, but has been hurt more by a loss of share – dropping from 94% in 2005 to 84% in 2019. We will look further, but Northville may be growing at Plymouth-Canton’s expense.
A generally rosy outlook for Rochester (slide 12). Population continues to rise with equally healthy gains in marketshare. A fly in the ointment is dip in school aged proportion, but this is probably overcome by schools of choice enrollment. EDIT: I don’t think this is the case. More to come in ensuing posts and research.
Alphabetically, Troy brings up the rear, but clearly this is a district gaining popularity, indicated by a huge increase in their Enrollment percentage of school aged population, especially from 2005 to 2019. Again, likely an indicator of school of choice success which is outpacing their slight decrease in proportion of school aged children. EDIT: This guess may have been more accurate as Troy’s school choice enrollment is substantial.
If you’re still awake, God Bless You! I hope this piece provides a glimpse for the complexity of data analysis on community and school demographics, finance, and policy. If someone suggests a simple answer or diagnosis of a district’s recipe for success (or struggles) I’d encourage keeping an open mind. Even though this entry contains a lot of data, it’s really the tip of the iceberg.
I look forward to examining the deeper dimensions of these icebergs in future posts.