Executive Summary
The impact on local schools is a crucial issue to consider when contemplating allowing the construction of incremental housing
The latest research shows that due to the state’s rapid decline in student population, enrollment levels do not track with new housing production. This is true for communities across the North Shore
Top school districts, rather than rejecting new construction, embrace it as a way to broaden their tax base and deliver excellent education outcomes for their students
—-------------------------------------------------------------------------------------------------------------------------
The impact on local schools is a crucial issue to consider when contemplating construction of new housing. This issue is particularly acute in Marblehead given the recent challenges the town has experienced with respect to school budgets, and ultimately the need to cut staff.
The logic advanced by opponents of 3A zoning in Marblehead is that building more housing will increase school enrollment, and those students will be a drain on the already stretched resources of our school system. This line of thinking makes several unproven assumptions and omits important facts.
The first crucial assumption to test is does constructing new housing result in increased school enrollment? The chart below compares the percent change in total enrollment on a full-time-equivalent (FTE) basis with the percent change in housing in a selection of North Shore communities between 2010 and 2022.
As you can see, in some towns with strong housing growth, such as Salem and Gloucester, their enrollment on an FTE basis has actually fallen. Marblehead's enrollment figures have also fallen, even though we have constructed relatively little housing compared to these other towns. In fact, the only town in the sample which saw an increase in enrollment was Beverly. This is, however, a small sample size and it is worth trying to understand how this dynamic plays out across the Commonwealth.
Fortunately, there are much more rigorous studies of the relationship between school enrollment and housing construction we can turn to. The Metropolitan Area Planning Council studied this exact issue in 2017 and updated their findings in 2024. Their key finding is that there is no relationship between housing construction and school enrollment.
The reason that housing construction and school enrollment have decoupled is the collapse in Massachusetts' overall school population. Between 2010 and 2020 the under 18 population statewide declined by nearly 17 percent. The result is that schools are generally underutilized and rather than housing construction being a predictor of enrollment, parental preferences play a larger role. And these preferences are more likely to depend on factors like proximity to good jobs or family networks than housing development.
The takeaway for Marblehead as it relates to 3A zoning is that we should not assume that the construction of new multifamily housing will actually increase enrollment.
Even if you reject the data and you are convinced that any new multifamily housing in Marblehead will necessarily increase enrollment, a critical factor to consider is timing. It is implausible to assume that the net new housing units (about 600 in total) permitted under a 3A proposal would all be built at the same time, or that the full complement would be built. The requirement for building permits and other town approvals means our school system would have plenty of time to assess the potential for any significant impact on enrollment from new development.
Furthermore, enrollment in Marblehead's schools is down by about 600 students from the 2013-2014 school year (see further down). It is unlikely that school capacity could have similarly collapsed in the same period.
On a separate note, the idea that an incremental student comes at the expense of the education of an existing student is too narrow a view and does not grapple with fixed costs.
To open a school in Marblehead, or anywhere, fixed operating costs must be paid. In this case, fixed costs are the expenses required to open the doors of a school building. The most obvious examples for Marblehead are utilities: Schools must be powered, heated and provided with water. It does not matter how many students are in the school, there will be some minimum charge to establish and maintain these connections. And schools have other fixed costs, at least one principal is required, at least one maintenance person and so on.
As a parent, what matters to you is the variable cost of educating your child. Variable costs in this case are spending items like teacher salaries and classroom supplies.(1) You (probably) don't care what the fixed costs are because those costs don't represent funds going directly to your child's education. Instead what you want to know is how many of the total dollars appropriated by the town are actually going to your child specifically?
Simply put, the fewer children in the school system, the larger the share of school funding that will have to go to fixed costs and the smaller the share of funds that will actually directly improve the education of your child. Let's use an example to illustrate.
Imagine a school district with two children where the per-student spend is $20,000 with fixed costs of $10,000. The variable cost / funding available for each individual student is $15,000. Now imagine a third child arrives in the district. Total spending rises to $60,000 but fixed costs remain fixed (it's in the name). So now the amount that is spent on each specific student in terms of teachers and programming rises to $16,667 ((3 20k per student) - $10k FC) / 3) = $16,667.(2)
You may object here that the town needs to find the extra $20,000 for the third student, and we will get there. But to drive this point home, imagine that Marblehead's student population continues to decline as it has for the past decade. If we got down to a single student, or even 10 students, would their education be richer and fuller thanks to the fact there are fewer students around to draw budget away from them?
We noted that towns have to bear the cost of educating incremental students enrolled in the district. If the cause of an incremental student's arrival is the construction of new housing, the value of that housing at least partially offsets the cost of educating the new student. Enter "new growth taxation".
New growth taxation is levied on new construction and significant renovations.(3) It allows towns to increase their total property tax receipts without having to increase the property tax rate on existing residents. By way of background, the total amount of property taxes collected by the town cannot increase by more than 2.5 percent from one year to the next - a consequence of Proposition 2 1/2. Given that inflation has exceeded 2.5 percent for the past several years, Proposition 2 1/2 has forced Marblehead to confront overrides to meet higher school costs, which have been contentious and so far unsuccessful.
New growth taxation, by contrast, adds to the town's overall property tax base incremental to Proposition 2 1/2 limits. In other words, new growth taxation is not subject to the 2.5 percent cap and allows the town to increase its property tax receipts faster than 2.5 percent without requiring an override. Also, crucially, it increases the tax base without increasing tax rates on existing residents.
So the basic math problem in respect of new multifamily housing and its impact on school funding becomes will the assessed value of the new growth taxation arising from new housing construction offset the cost of educating new students?
Given that housing production is not a reliable proxy for student enrollment changes, as we highlighted earlier, this is a difficult problem to solve with forecasting. An alternative approach is observation. We can ask to what extent do top school districts across Massachusetts rely on new growth taxation?
In theory, if new students entering a district through new housing were harmful to the education of existing students, we would expect to see well regarded school districts shying away from too much new construction and the associated new growth taxation that comes with it.
To get a sense for this, we compared Marblehead to ten school districts ranked among the best in the Commonwealth by new growth taxation as a share of their prior year property tax receipts. We ran this comparison for FY2020-FY2024 and we also included the statewide average, weighted by the base year's total tax receipts.
Marblehead is notable for coming in last with respect to the contribution of new growth to the town's finances and thus its capacity to fund an excellent school system for its children. Hopkinton, whose public schools rank as the best in the Commonwealth, had more than four times the contribution to 2024 property tax receipts from new growth compared to Marblehead. In fact, every town in the sample had a higher contribution of new growth taxation in 2024.
But we can go further with this data. By combining pupil enrollment figures and changes in enrollment from year to year, we can estimate how much we can expect new growth to contribute to offsetting the cost of new students in a district on a per student basis (note that these figures are whole pupil enrollment rather than FTE).
To do this, we first adjust the amount of new growth taxation available for the upcoming fiscal year to account for the fact that not all new growth taxation funds would be available for the exclusive use of schools. In Marblehead's 2024 budget, approximately 42.3 percent of funds were spent on schools. Therefore we reduce all the new growth dollar figures by that amount. We then exclude years with declines in enrollment (because this leads to nonsensical values), and we remove Marblehead itself to focus on outcomes among top school districts.
The results of this analysis are below. As you can see we can discard the maximum in the dataset as an outlier.(4) But the rest of the results cluster reasonably tightly with the average in the sample being about $18,000 and the median coming in just below $17,000. For comparison, Marblehead's per student cost for the 2023-2024 school year was about $18,000.
However, it is true that over time the total enrollment across Marblehead's public schools might rise as a consequence of new housing, and in about half of cases in our analysis above, new growth taxation comes in below the estimated cost to educate a student in Marblehead. This is a concerning outcome for some, and a positively apocalyptic one if the campaign against 3A zoning is to be believed. So if more students are bad, fewer students must be better, right?
The chart below shows Marblehead's public school enrollment on a whole pupil basis over the past decade.
As you can see, even before the significant decline in enrollment caused by COVID-19 the district's total enrollment figures were on a rapid downward trend.
The question this chart anyone concerned about school funding ought to ask is has it gotten easier to fund our schools over the past decade, or harder? Consider the current challenges the town faces in its collective bargaining process with the education union. Would that negotiation be easier if Marblehead only had a thousand students? Five hundred? What if we had just one student left?
Instead of seeing the potential (but not certain) addition of new students to the public school system as a result of new housing as a threat to the education of existing students, they instead should be seen as broadening the base of support for funding the education of all the town's students. This appears to be the attitude taken by top school districts across the Commonwealth, and in the spirit of education, there might be something to learn from that.
(1) Teacher salaries effectively operate with a "breakpoint". If one additional student enters a school district, it is not possible to hire 0.01 of an extra teacher (or whatever the right increment would be relative to prescribed class sizes), but across a student population as large as Marblehead's, teachers can effectively be viewed as a variable cost.
(2) Note that it doesn't matter what number you use for fixed costs here, the result of a higher proportion of variable cost being available for the third student will always be the same.
(3) While we do not have access to data splitting new growth taxation that comes from new construction vs. renovations, we would expect the latter to mostly be the result of choices made by incumbent owners. I.e., generally renovation-related new growth would not be related to a potential change in school enrollment totals.
(4) Shrewsbury saw an increased enrollment of seven students in the 2022-2023 school year and new growth taxation dollars of $1,316,835 which created the outlier here.
Comments