Grosse Pointe
in Focus

Michigan Benchmark is a nonpartisan public data organization. This analysis draws on fifteen years of student enrollment records, staffing data, financial reports, community demographic data, and property tax records. All data sources are described in the appendix. Our findings are reported as the evidence shows them, including where the data complicates simpler narratives and where it raises questions that warrant further investigation.

Between 2010 and 2026, GPPSS lost 2,021 students in total enrollment. That number, reported in isolation, suggests a district contracting uniformly across its population. The disaggregated data tells a different story.

The district's economically disadvantaged population (students qualifying for assistance based on household income) did not decline. It grew. GPPSS enrolled 871 economically disadvantaged students in 2010 and 1,083 in 2026, an increase of 212 students at a time when total enrollment fell by more than two thousand. The entire net enrollment decline, and approximately 200 students beyond it, came from the non-economically-disadvantaged population. Students classified as economically disadvantaged stayed. Students outside that classification left.

This pattern is not unique to GPPSS. All six comparable peer districts show rising economically disadvantaged percentages over the study window. But the rate of increase at GPPSS is among the highest in the peer group, exceeded only by Birmingham Public Schools, which shares a similar demographic profile and private school competitive environment.

One methodological note: Michigan changed its methodology for classifying economically disadvantaged students around 2018, transitioning to a direct certification model based on SNAP, Medicaid, and other assistance programs. The jump in the economically disadvantaged percentage from 13.7 to 18.2 percent between 2017 and 2018 reflects in part this reclassification. The structural direction of the trend is real and consistent across the methodology change: a rising economically disadvantaged share in a declining enrollment district. Year-to-year comparisons spanning 2017–2018 should be interpreted with this caveat in mind. Racial composition data from the same source is consistent with the economic pattern but does not alter the finding.

The most consequential and least-reported finding in this analysis concerns not where the district has been, but where it is now. After more than a decade of structural enrollment decline, GPPSS's entering kindergarten class now roughly equals its graduating senior class for the first time in at least fifteen years. The engine of decline has stopped running.

In 2010, GPPSS graduated 777 seniors while enrolling only 412 kindergarteners, a structural gap of 365 students. Every year that gap persisted, total enrollment fell by the size of the difference. At its peak the gap exceeded 300 students annually. By 2022 it had narrowed to roughly 100. In 2025 it inverted: for the first time, the entering class was larger than the graduating class.

The rate of total enrollment decline tells the same story in different terms. GPPSS lost approximately 530 students in FY2021, 307 in FY2022, 126 in FY2023, and fewer than 50 in each of the two years since. The annual decline in FY2026 is in single digits.

This finding has direct implications for enrollment projection work. Models that project continued steep decline at historical rates are using an input assumption the data no longer supports: that the demographic forces driving the decline are ongoing. The structural rebalancing has largely occurred. The question for the district going forward is not how many more students it will lose, but whether it can maintain its current enrollment level and serve a student body of different composition than it was built for.

The enrollment trends documented above eventually produced a physical consequence. In June 2019, the GPPSS board of education voted to close two elementary schools: Charles A. Poupard Elementary in Harper Woods and Robert Trombly Elementary in Grosse Pointe Park, effective at the end of the 2019-20 school year. The board simultaneously reconfigured the district's grade structure, moving fifth grade from elementary schools to middle schools. Trombly's students were consolidated primarily into George Defer Elementary. Poupard's students were distributed between Stevens T. Mason Elementary and John Monteith Elementary.

The building-level enrollment data from 2010 through 2026 allows the full system to be examined, including the two buildings that closed, without relying on the narratives that surrounded the decision at the time. What the numbers show is both more precise and more complicated than the headline enrollment decline story.

Three findings from the building-level data warrant specific attention.

Neither closed building was in acute enrollment distress in the year of the vote. Poupard enrolled 304 students in FY2019, roughly the same as in FY2017 (303) and FY2018 (306). Trombly enrolled 248 in FY2019, down modestly from 267 in FY2017 and 257 in FY2018. Both schools were small, the two smallest elementary buildings in the system, but neither was in freefall. By contrast, Defer and Richard had declined more steeply in the years prior: Defer had fallen from 414 in 2010 to 307 in 2015 before recovering. The buildings selected for closure were the smallest, not necessarily those showing the most acute recent decline.

The reconfiguration coincided with a sustained acceleration in School of Choice departures that has not reversed. GPPSS is one of a small number of Michigan districts that does not participate in School of Choice and accepts no inbound transfer students. The outflow side of that equation tells a story the enrollment totals alone do not capture.

Two currently operating buildings now enroll fewer students than the schools that were closed. Lewis Maire enrolled 244 students in FY2026, and has been below 260 since FY2022. Trombly closed with 230 students in its final year. Kerby enrolled 299 in FY2026, down from 363 in 2019, a decline of 18 percent in seven years. Neither building has stabilized. The enrollment threshold that produced the 2019 closure decision, if applied uniformly today, would identify Maire as the district's most at-risk elementary building.

School enrollment follows community demographics with a lag of roughly five years, the interval between household formation and kindergarten entry. To understand where GPPSS enrollment is headed, the right place to look is not the district's own historical trend but the age structure and household composition of the communities it serves.

The Grosse Pointe communities are aging. The median age of the district's service area has risen from approximately 42 to nearly 45 over the study window. More than one in four households now receives retirement income, up from 22 percent in 2010. The share of the population composed of school-age children has fallen from 26.3 to 23.4 percent.

The contrast with Novi Community Schools is the clearest illustration of what community age structure means for school enrollment. Novi and GPPSS are comparable on household income; both communities have median household incomes in the $105,000–$115,000 range. But Novi's retirement income household rate is 17.5 percent versus GPPSS's 26.5 percent, a nine percentage point gap that represents 51 percent more retirement-income households proportionally. Novi's median community age is 40.2 versus GPPSS's 44.7. Novi's share of children in the population is 25.4 percent versus 23.4 percent.

The result: Novi's enrollment has grown from 6,507 to 6,781 over the period when GPPSS declined by 2,000 students. Novi captures 97 percent of the school-age children in its communities. GPPSS captures 86 percent.

The housing stock comparison helps explain why this gap exists and why it is structural rather than cyclical. The median Grosse Pointe home was built in 1951, more than three decades before the median home in Novi or Northville. Older, established neighborhoods attract a different household profile than communities where housing stock is newer: established families, empty nesters, and retirees rather than young couples in the household formation years. Nearly one in five Grosse Pointe homes was built before 1940. This is not a condition that changes quickly. The community's physical character shapes the demographic profile of the families who choose to live there, and that profile produces fewer school-age children per housing unit than the communities with which GPPSS competes for enrollment.

The in-migration data reinforces this. In any given year, 6.5 percent of the Grosse Pointe population arrived from a different city, below Birmingham at 8.4 percent and Bloomfield Hills at 7.7 percent. Lower in-migration means the natural replenishment of young families that drives enrollment stability in other communities is slower in Grosse Pointe. The pipeline from outside is constrained by the same housing stock dynamics that shape the resident population.

A third structural headwind compounds the housing and demographic picture: property tax burden. Grosse Pointe homeowners face the highest total homestead millage in the peer group. The premium has nothing to do with the school district.

The breakdown makes the finding precise: the school district's homestead levy is identical to Troy's: 9.63 mills in both communities. Every mill of the premium comes from outside the district's control. Wayne County levies 7.71 mills compared to Oakland County's 5.52, a 2.19-mill difference. The Wayne RESA levy of 5.38 mills exceeds the Oakland ISD levy of 3.13 mills by 2.25 mills. Municipal levies within the five Grosse Pointe communities range from 16.9 to 19.8 mills, between 7 and 10 mills above Troy's 9.79. None of these are school district decisions.

Municipal levy variation within the district is meaningful and worth acknowledging directly. The five Grosse Pointe communities: Grosse Pointe city, Farms, Park, Woods, and Shores, carry total homestead millage ranging from approximately 45.6 to 48.5 mills, a relatively tight spread. Harper Woods, also within the GPPSS service area, carries 62.4 total homestead mills driven by its 33.7-mill municipal levy, a substantially different burden that reflects that community's own fiscal structure and should not be conflated with the Grosse Pointe proper communities in a housing comparison.

For a family choosing between a home in a representative Grosse Pointe community and a comparable home in Troy, the millage-rate premium is substantial: 13.76 mills at Grosse Pointe Park's representative rate of 47.84 mills. Translated into a modeled homeowner burden estimate, Michigan Benchmark applies this premium to an estimated taxable-value base rather than treating the ACS market-value figure as the tax base. Using a taxable-value base of approximately $152,550 (roughly 50% of the ACS median owner-occupied home value of $305,100), the taxable-value-adjusted estimate for the GP Park–to–Troy premium is approximately $2,099 per year, or roughly $175 per month. The five GP communities bracket the estimated annual homeowner burden between approximately $1,753 (GP city, the lowest-mill community) and $2,201 (GP Woods, the highest). These are modeled homeowner burden estimates, not parcel-level tax bills; actual tax exposure depends on each property's taxable value and assessment history. The rate comparison remains the most defensible measure. Even on a taxable-value-adjusted basis, the premium is material, and it accrues entirely from the jurisdictional geography of Wayne County rather than from anything the district has done or can undo.

This tax premium interacts directly with the housing stock and in-migration constraints already documented. A community that already struggles to attract young family buyers because of older housing stock faces an additional cost headwind with every comparable purchase against Oakland County alternatives. The three structural constraints reinforce one another: aging housing stock, high retirement income household concentration, and the Wayne County tax premium. None are within the district's control. Together they explain why the in-migration rate that would organically replenish the school-age population has been running below peers for over a decade.

The broad enrollment decline masks a more specific behavioral pattern that the cohort-level data reveals. Tracking individual entering classes forward through the grade levels produces a finding that runs counter to the conventional narrative about families abandoning GPPSS: students who enroll in Grosse Pointe Public Schools tend to stay, and more join them along the way.

The kindergarten class that entered GPPSS in 2012 had 477 students. By grade 12 in 2024 that cohort had grown to 586, a gain of 109 students over twelve years. The 2016 entering class of 414 had grown to 491 by grade 9, already 19 percent larger than it started. Cohort growth, not cohort attrition, is the pattern.

Within that growth pattern, one anomaly is consistent across every tracked entering class: enrollment recovers sharply at grade 9, gaining back students who had left during the middle school years and in most cohorts exceeding prior levels. This grade 9 return is the most consistent and cross-cohort-verified signal in the data.

Comparing the two cohorts clarifies what the data does and does not establish. The grade 6–7 dip visible in the 2015 cohort coincided with the first years of the building reconfiguration and COVID disruption, both of which independently drove families out of the district during those years. That specific dip is not cleanly replicated in the 2016 cohort, where grade 6 and 7 are essentially flat. What is replicated, with greater consistency, is the grade 9 return: the 2015 cohort gained +41 at grade 9, the 2016 cohort gained +48, and the same pattern holds in the 2012 and 2017 cohorts at +21 and +34 respectively.

The most plausible interpretation is that families who leave GPPSS during the middle school years, whether for private school, relocation, or any other reason, return at a meaningfully higher rate at grade 9 than the district's overall retention pattern would predict. The high school is drawing students back. That is the durable behavioral signal in this data, and it holds across pre- and post-reconfiguration cohorts alike.

A secondary structural note: the 2019–20 reconfiguration moved grade 5 from elementary to middle schools. For cohorts that have progressed through the system since then, the first-year middle school transition now occurs at grade 5, not grade 6. The data shows modest softness at grade 5 in post-reconfiguration cohorts, a pattern worth monitoring as more cohorts complete the full grade 5–12 arc.

The peer comparison sharpens the finding considerably. GPPSS is the only district in the group where the grade 9 return index exceeds 1.0 in every year of the study window, a consistency no other peer district matches. Its six-year average of 1.11 is the highest in the group. Novi, which faces less private school competitive pressure and maintains more stable enrollment, averages 0.97. Grade 9 is actually slightly smaller than grade 8 on average there.

Whether the grade 9 return reflects a stronger high school offering, a weaker middle school competitive position, or both operating simultaneously is a question the enrollment data alone cannot resolve. What the data does establish is that the pattern is specific to GPPSS in its consistency and magnitude, most pronounced in the districts where private school competition for middle school students is most active. Bloomfield Hills, which sits adjacent to Cranbrook Kingswood, showed a similarly elevated return index in 2021–2023 before moderating. The districts least affected by private school competition, including Novi and to a lesser degree Birmingham, show the weakest or most volatile return signal.

Michigan's foundation allowance funding model is heavily enrollment-driven. Even where hold-harmless provisions and local revenue mechanisms complicate the exact dollar effect in a given year, enrollment loss creates recurring fiscal pressure: a smaller enrollment base weakens the operating revenue foundation over time. The enrollment decline documented in this report is therefore not an abstract demographic trend. It is a sustained pressure on the district's revenue structure that has been accumulating for fifteen years.

But the enrollment decline at GPPSS has not been uniform across the student population. As documented in Finding 1, the students who have left are disproportionately general education students not classified as economically disadvantaged. The fiscal issue is not the worth of any student group; it is that enrollment-driven revenue can weaken faster than fixed and legally mandated service obligations adjust. Students with Individualized Education Programs carry legally required service obligations that do not diminish with district size, enrollment level, or revenue. The combination produces a structural squeeze that proportional figures make visible more clearly than any dollar amount.

The ratio of special education staff to classroom teachers is the most direct measure of how the district's instructional footprint has shifted. In FY2010, for every 100 classroom teachers GPPSS employed, there were 49 special education staff — instructional aides, certified special education teachers, consultants, and support personnel combined. By FY2025 that ratio had risen to 64.5. By FY2024 it peaked at 67.9. The special education staffing footprint, held essentially flat at approximately 205 FTE throughout the study window, now represents a substantially larger share of the district's total instructional workforce. This is not because special education hiring grew; it is because the teacher count declined as general enrollment fell.

This is not a staffing policy problem. It is the mechanical consequence of a fixed legal obligation meeting a shrinking revenue base. Federal law requires districts to provide specific services to every student with an Individualized Education Program. Those obligations are calibrated to the number and needs of special education students, not to total enrollment or total revenue. As the general education enrollment base contracts, the district's enrollment-driven revenue weakens, while the special education staffing obligation does not move.

The general education share of enrollment fell from 89.7 percent in FY2010 to a low of 78.6 percent in FY2023, before recovering modestly to 84.7 percent in FY2025, a recovery that reflects both some economically disadvantaged reclassification and a stabilizing total enrollment. Students who generate foundation allowance revenue without mandatory add-on service costs have shrunk as a share of the student body throughout the decline. The district is now serving a student body with a more complex operating profile — one that carries higher mandated service obligations relative to total enrollment — than the one it was designed and sized for.

A full accounting of the financial consequence of these structural shifts requires analysis of real per-pupil revenue adjusted for inflation, pension cost dynamics under the Michigan Public School Employees Retirement System, and the specific revenue mechanisms available to GPPSS under Michigan's funding structure. That analysis is beyond the scope of this report. What this report establishes is the structural precondition: the composition of the student body has shifted in ways that increase per-pupil service cost while the enrollment-driven revenue mechanism contracts. Those two vectors are moving in opposite directions, and the proportional measures documented here show that the gap has been widening for fifteen years.

The five findings in this report are not independent observations. They describe a system under a specific and compounding kind of pressure. Understanding that pressure as a system, rather than as a list of problems, is the most important thing this evidence makes possible.

GPPSS is not a district in freefall. The pipeline has stabilized. The entering kindergarten class now exceeds the graduating senior class for the first time in fifteen years. The structural engine of enrollment decline has, for now, stopped running. That is a meaningful and underreported fact, and it matters for how the district frames its own situation going forward.

But stabilization is not recovery, and the operating baseline of what has stabilized is different from what the district was built to serve. The students who left were disproportionately students not classified as economically disadvantaged — the portion of the student body whose enrollment drives recurring operating revenue without generating mandatory add-on service obligations. This is not a judgment about students or families; it is an operating-finance point about enrollment mix and fixed costs. The students who stayed, along with the students now entering, include a growing share of special education students with legally mandated service obligations that do not diminish with district size. For fifteen years, enrollment decline and the rising share of mandated service obligations moved in opposite directions, creating a widening structural gap. Stabilization closes the gap on one side: the annual revenue pressure from enrollment loss has eased. But the district has stabilized at a smaller scale and a more complex operating baseline than it was designed for. That is the new operating reality. The financial implications of it have not resolved; they have simply stopped getting worse.

The community aging data establishes why demographic recovery through natural replenishment will be slow. The Grosse Pointe communities have among the oldest housing stock and highest retirement income household rates in the peer group. The five Grosse Pointe communities carry homestead millage ranging from 45.6 to 48.5 mills, between 11 and 14 mills above comparable Oakland County communities, driven by county, RESA, and municipal levies rather than school district decisions. At a representative community rate, the taxable-value-adjusted homeowner burden estimate for the GP Park–to–Troy premium is roughly $2,100 per year, based on applying the millage gap to an estimated taxable-value base of approximately 50% of the ACS median owner-occupied home value; this is a modeled homeowner burden estimate, not a parcel-level tax bill, and actual exposure varies by property. These are structural conditions, not temporary ones. The families most likely to generate school-age children are not arriving at the rate that would rebuild enrollment through organic community renewal. The district cannot wait for the community to regenerate itself on a timeline that matches its fiscal needs.

Against that structural backdrop, the grade 9 return finding carries specific weight. GPPSS has the most consistent and largest grade 9 return effect in the peer group, a pattern that holds across four cohorts and predates both COVID and the reconfiguration. Whatever its cause, the district's high schools are drawing students back. That is a competitive asset of real value in a system where demographic headwinds limit what natural replenishment can provide. Protecting and understanding it is a strategic priority the data specifically identifies.

The School of Choice outflow is the finding that most directly connects a policy decision to an enrollment outcome. The annual departure rate tripled in the years following the building reconfiguration, moving from a range of 18 to 69 students per year before 2019 to a sustained level of 100 to 166 afterward. The data cannot assign causation: the reconfiguration and COVID overlapped. But the outflow channel opened wider during that period and has not closed. A district that accepts no inbound transfers and loses 166 students per year outbound is operating with a structural competitive asymmetry that compounds annually. Every neighboring district can draw from GPPSS. GPPSS cannot draw from them.

Taken together, the evidence in this report points toward four questions that the district's leadership, community stakeholders, and partners like PFAE are positioned to act on, not as prescriptions, but as the specific inquiries the data most directly opens.

What is happening in the middle school years, and what can be done about it? The grade 9 return is consistent enough, and distinctive enough to GPPSS, that it warrants a dedicated investigation. Survey data, private school enrollment trends, and family exit interviews would allow the district to distinguish between a competitive vulnerability and a structural preference, and to target any response accordingly. The data identifies the question with precision. Answering it requires going beyond what enrollment counts alone can show.

What does the School of Choice outflow represent, and is it recoverable? The acceleration after 2019 may reflect a COVID-era behavioral shift that has now become permanent, or it may reflect a response to the reconfiguration that is still resolving. Understanding which families are leaving, to which districts, and for what reasons would allow the district to distinguish between displacement (families who left because of a specific decision) and structural competition that would exist regardless. The two require different responses.

How does the financial structure of the district interact with the enrollment and demographic shifts this report documents? The structural squeeze in Finding 5 establishes the precondition. A full analysis of per-pupil revenue adjusted for inflation, MPSERS pension cost dynamics, and the hold harmless mechanisms embedded in GPPSS's funding structure is needed to quantify the financial consequence. Michigan Benchmark intends to publish that analysis as a companion to this report. The enrollment and demographic evidence here provides the foundation.

How should enrollment projections be recalibrated? The pipeline stabilization finding has direct implications for any projection model currently in use. Models that project continued steep decline are using an input assumption the data no longer supports. At the same time, models that assume demographic recovery through natural community regeneration may be overstating what the housing stock and in-migration data suggests is feasible. The right projection is probably neither: a stabilized but compositionally different district, operating under structural financial pressure, with specific vulnerabilities in the middle school years and specific strengths at the high school level.

All data in this report is drawn from public sources. Michigan Benchmark maintains a structured analytical warehouse integrating these sources for longitudinal analysis. Sources are referenced throughout this report by their public-facing institutional names.