I tuned in to last Monday’s School Committee meeting, eager to hear discussions around enrollment capacity and the “Mass Development Readiness Meeting in preparation for Super Town Meeting.”
In an effort to guide the committee and community as to whether Harvard’s schools will have capacity to absorb the projected influx of school-aged kids from Trinity Financial’s proposed Vicksburg Square redevelopment, interim superintendent Joe Connelly prepared an enrollment forecast using the “Cohort Survival Methodology.”
—Virginia Justicz
This is the same methodology that the New England School Development Council (NESDEC) uses to project future enrollment for Harvard (and every other district in Massachusetts) on an annual basis. Cohort survival analysis looks at past movement of age groups and tracks enrollment growth or decline as they move from grade to grade, often with predictable patterns. This is referred to as their “survival rate” in the school system from birth to kindergarten, and then for each subsequent year.
For Harvard, where school-age child numbers are relatively small, the model is sensitive to small fluctuations in assumptions. For instance, a slightly larger-than-expected ratio of kids who enroll in kindergarten to the number of births in Harvard five years prior can throw the projected enrollments off significantly.
The analysis itself is useful as a starting point for discussion about what trends might occur in the future.
What is unsettling was the context (or lack thereof) with which Dr. Connelly presented these projections and used them to conclude that Harvard’s enrollment will decline over the next 10 years.
Within a matter of minutes, most of the committee was nodding along, satisfied that there will indeed be capacity for any additional school-aged kids who would reside in the proposed 246-unit Vicksburg development. I did not hear anyone question the confidence levels of the assumptions in the model, or discuss what happens to enrollment if the economy rebounds and other housing developments in Harvard and Devens come to fruition.
Before the recession, many fretted that school enrollment would soon exceed capacity because of new development and “empty nest” housing turnover. Then, as building permits and housing sales dropped precipitously, we began to project declines. This is the problem with using recent history to predict future enrollment. Past enrollment patterns do not include that perfect crystal ball for the unpredictable macro trends that will drive changes in the future.
The key assumption in Dr. Connelly’s analysis is that we will continue to see declining incoming kindergarten classes since birth rates in Harvard have dropped. In the model, the small kindergarten group then progresses through the grades, adding or losing a percentage each year based on historical trends.
Historically, however, birth rates have not been very good predictors of eventual kindergarten enrollment in Harvard. Many families move to Harvard after their children are born, and the timing of in-migration depends on many factors, including employment, relative housing availability, and school desirability. Furthermore, even when a kindergarten class is small (e.g. 60), organic growth (people moving into town) can easily expand a kindergarten class of 65 to a third grade class of 88 or 90 several years later.
The conclusion that we could handle a large influx of Vicksburg kids (even in the unlikely event that they are spread evenly across the K-12 grades) fails to consider additional scenarios that could put significant pressure on existing capacity:
- Future residential construction in Harvard
Several years ago, committees in town were looking at a handful of 40B projects in Harvard’s pipeline, including a large Shaker Hills development, that could have added many school-aged kids if the economy had not put a halt to new development. It is possible that an economic recovery would once again ignite future building activity. There is still a significant amount of buildable land.
- Other residential construction at Devens
The NESDEC long range planning report (delivered to the School Committee in April 2011) states: “Exclusive of Vicksburg Square, 282 housing units have been approved at Devens. 176 are remaining to be built. In addition to the 20 ‘net zero energy’ housing units to be constructed at Devens, there are 156 remaining units that would likely be constructed by 2020, adding an additional influx of school-aged kids.”
- Housing turnover as empty-nesters choose to leave Harvard
The NESDEC report also includes the following: “Despite the current conditions in the residential housing market, there are factors which indicate Harvard is likely to experience significant housing turnover in the near future. According to realtors, the Baby Boom cohort (which will account for 30% or more of the Harvard population by the latter part of the decade) is likely to downsize to smaller homes once the housing market rebounds from its present slump. These people would be vacating three and four bedroom homes that would quite likely be inhabited by new families with school-aged children.”
This has the potential to blow the more optimistic projections out of the water and serve to overload our schools.
I am discouraged that the superintendent and School Committee would be so quick to use the cohort survival analysis in a vacuum to provide a simple answer to a critical question that will impact our schools five to 10 years from now.
Even NESDEC, which uses this model on a regular basis, points out that, while it is a good predictor in the near term, it is not particularly accurate when looking out over longer time frames. Cohort survival analysis isn’t a bad methodology, but it relies heavily on the recent past and cannot predict future trends that will change patterns.
Most importantly, it is certainly not robust enough to use as the basis for a very important decision about whether or not our schools can absorb a fully occupied Vicksburg. At the very least, we need to consider the viable scenarios that could alter these enrollment projections and present the risk of large class sizes, overcrowding, and a diminished educational experience if the numbers trend in a different direction.
Virginia Justicz, of Woodside Road, was a member of the School Committee from 2007 to 2011. She is a parent of three Hildreth Elementary School students.
