The largest school district in the United States has 1,213 schools and 976,771 students.
-National Center for Education Statistics, 2017
Not surprisingly, New York City has the largest school district in the country, with more than 1,200 schools and almost one million students served in 2017. Los Angeles Unified School District, the second largest school district, has only two-thirds the number of students (approximately 621,000) and yet still has over 1,000 school buildings within its boundaries. By contrast, the smallest school districts on record in the United States has fewer than 100 students, and a few of these exist in almost all states (NCES, 2017). The average school district has between five and six schools (NCES, 2017) and the average school district in 2019-2020 was 526 students (Research.com, 2022). There are huge disparities in the numbers between large and small districts, and this has a significant impact on students and the surrounding community. Those who study education and speculate on the ideal size of school districts cite geographic constraints, financial costs and benefits, community obligations, and more in what dictates the size of a district. Ultimately, how all of these factors impact student performance, and the success of students overall should drive the ideal size of school districts.
Introduction
My personal experience with school district size
| Attending School | I grew up in a district of about 1,200 students, K-12. One building, universal busing (both before/after school and after extracurriculars), teachers who mostly specialized in grade level instruction, athletic programming that did not necessitate tryouts, and extracurricular activities supported by a local community were predominant elements of my experience as a student. |
| Teaching School | My first full-time teaching job came in a rural community close to where I grew up, in a school district of about 1,500 students. Another district comprised of one building, with all of the other elements in common with my experience as a student. The exception to this was the lack of universal busing, as the geographic location of the district was centered within a small town, whereby students living within one mile (appropriate to their ages) were ineligible for busing. |
| Leading School | As a building principal, I was part of a school district that is the largest school district in New York State with regard to land size, with over 250 square miles within its boundaries. At the time, the district had seven schools with about 4,500 students. This district utilizes universal busing, and though the elementary schools have teachers relegated to grade-level instruction, the middle and high schools have multiple teachers assigned to grade levels. Tryouts for athletics exist, but so does an athletic trainer and assistant superintendents. A supportive community is still predominant. |
| Leading School District | In my current role as assistant superintendent, I am in a school district of about 4,300 students, Pre-K-12, with eight buildings and an outpost building that houses some administrative roles. A small city school district, not all students are eligible for buses, intramural sports are there for those who try out for teams and are not successful, assistant superintendents and other administrators support building principals, and there are additional community schools’ elements (dental and health clinics, along with counseling services and after-school programming) are present for student and family support. |
Ideal District Size – Geography
While public school enrollment has increased in the last eighty years, the number of public school districts has significantly decreased in the same span of time. The California Policy Center (2017) cites statistics from the National Center for Education Statistics that show there are a tenth of the public school districts in existence today than there were in the end of the 1930’s (13,672 in 2014-2015 versus 117,108 in 1939-1940). Geographically, schools have become much larger, stretching their boundaries upwards of hundreds of square miles, resulting in bus rides for students exceeding one hour each way. This is double the length of the average adult worker’s commute, last surveyed at 27.6 minutes in 2020 (U.S. Census Bureau).
In a study conducted by Cordes, Rick and Schwartz (2022) on the effects of long bus rides on student outcomes, the results indicated that the longer the commute, the more likely there was a small but negative impact on student performance. Significantly, however, the longer the bus ride a student has, the more likely the student was to experience chronic absenteeism. The caveat to Cordes et. al work is the impact of a longer bus ride on a student attending a school of choice. That said, 93 percent of students attending public school attend traditional public schools, which means that the majority of students do not experience the positive benefits of a longer bus ride stemming from attendance at a preferential school. Most students attend a school based on their attendance zone/boundaries, and, as such, would be more apt to see the negative impact of long busing on their academic performance.
Further, the time spent on a bus ride is not the only element that is part of student’s daily commute to and from school. Many students have ten to fifteen minute walks to bus stops and/or a wait for the bus to arrive in the morning. Add to this the early start times of some schools, which impact the sleep schedules of students, and a lengthy bus ride, particularly to school, could have significant albeit indirect influence on student performance.
Some believe that long bus rides could be replaced with virtual learning mechanisms for students, but this, too, comes with unintended consequences. For many students, online learning comes at a cost of academic performance; speculation since the COVID-19 shutdowns have alluded to decreased academic outcomes and fewer course completions for students. Add that to the isolation that is experienced by many students participating in online learning, and this is not necessarily an optimal solution to the problem of lengthy bus rides for larger sized districts geographically.
The ideal district size in consideration of geographical boundaries is one that has a bus ride that is at least comparable to that of an adult’s commute to/from work. This would substantially increase the number of school districts in existence, and other factors need to be examined in conjunction with this, but, in isolation, districts with boundaries that allow for approximately a 25-minute maximum bus ride seems to be appropriately ideal.
Ideal District Size – Fiscal Obligations
The first argument many jump to when discussing ideal district size is that of financial and the cost to run a small district. In practicality, it is likely that there is a small increase in cost for smaller districts to operate. This is likely because the maintenance and operations of buildings are relatively static across districts. A building of a particular size costs a similar amount to maintain and run, regardless of the human capital passing through its halls each day. Heating, electric, and general upkeep all are necessities of a building, no matter the number of people utilizing a space.
Aside from the operational costs, though, staffing and materials is really what drives the cost up for districts. Personnel costs (inclusive of salaries and benefits) often account for upwards of 80% of a district’s budget (Education Week, 2011). When thinking about the structure and needs of smaller versus larger school districts, staffing obviously plays a distinct role. The number of staff needed to maintain a small staff to student ratio increases proportionally with the increase in students enrolled. But this is somewhat misleading. With the increase in student enrollment comes the need for greater management of procedures, requirements, and other administrative components. There may also be an increased need for support staff, given the context of the district’s setting and the student population. A smaller district may have great needs, but might be able to meet these needs with community partners or a smaller number of staff. While larger districts can certainly have community partners, they often cannot partner to the extent needed to offset the number of students exhibiting needs for that particular service.
An example: A small school district might be able to acquire .5 FTE of a community partner’s staff at a shared cost to offset the needs of some students out of their population of 650-2,400, but a larger school district of 13,000-39,000 students would need between 16-60 FTE in order to offset that same proportional need within its student population. This is not possible for most community organizations who wish to support their local school districts and student needs.
Examining the hypothetical and very basic chart below, one can see that the trends indicated by research that cost savings are maximized in settings that are small to medium sized, and the savings are minimized once a “threshold of approximately 6,000 students is reached” (Schiltz & De Witte, 2017, p.1051).
| District Expense | Large School District (39,000 students) | Medium School District (13,000 students) | Small School District (2,400 students) | Extremely Small School District (650 students) |
| Average Cohort K-12 | Approximately 3,000 students | Approximately 1,000 students | Approximately 200 students | Approximately 50 students |
| Building Maintenance/ Operations | $1,000,000 x 30 schools = $30,000,000 | $1,000,000 x 10 schools = $10,000,000 | $1,000,000 x 3 schools = $3,000,000 | $1,000,000 x 1 building = $1,000,000 |
| Administrative Overhead | 30 principals x $90,000/year = $2,700,000 20 assistant principals x $75,000/year = $1,500,000 10 directors x $110,000/year = $1,100,000 3 assistant superintendents x $130,000/year = $390,000 1 superintendent x $225,000/year = $225,000 | 10 principals x $90,000/year = $900,000 5 assistant principals x $75,000/year = $450,000 5 directors x $110,000/year = $550,000 1 superintendent x $175,000/year = $175,000 | 3 principals x $90,000/year = $180,000 2 assistant principals x $75,000 = $150,000 1 superintendent x $150,000/year = $150,000 | 1 principal x $90,000/year = $90,000 1 superintendent x $125,000/year = $125,000 |
| Staffing (~20 students/ class) | 1,950 x $50,000/year = $97,500,000 | 650 x $50,000/year = $32,500,000 | 120 x $50,000/year = $6,000,000 | 33 x $50,000/year = $1,650,000 |
| Curriculum/ Materials Purchases | $2,000/student = $78,000,000 | $2,000/student = $26,000,000 | $2,000/student = $4,800,000 | $2,000/student = $1,300,000 |
| Busing | $1,000/student = $39,000,000 | $1,000/student = $13,000,000 | $1,000/student = $2,400,000 | $1,000/student = $650,000 |
| Other Extraneous Purchases | $100/student = $3,900,000 | $100/student = $1,300,000 | $100/student = $240,000 | $100/student = $65,000 |
| TOTAL | $254,315,000 | $84,875,000 | $16,920,000 | $4,880,000 |
| Per Pupil Expenditure | $6,520/pupil | $6,528/pupil | $7,050/pupil | $7,507/student |
For a difference of $522 per student each year, a district can exist in a small community with perhaps a smaller geographic profile and a system of support that is in close proximity to its student population. A district with 2,400 students would have an added cost of about $1,252,800 as opposed to a larger district. This might seem implausible, but this is a small cost when considering the other benefits to the students and the community-at-large. Related to this, Schiltz and De Witte (2017) emphasize that “focus should be shifted towards long-run effects of district consolidations on costs, and most importantly student outcomes” (p. 1063).
Ideal District Size – Student Performance
Perhaps the most important component of identifying an ideal district size is examining performance of students within a given size range. Of course, there will be outliers in any district, as high performance and low performance is indicated in every setting. But, if one adheres to the bell curve and then examines the function of proficient performance, it is clear that smaller districts do tend to outperform larger districts.
Using the table below, with the five of the largest districts in New York State (outside of New York City), five small city school districts, five medium-sized districts, and five small, rural districts throughout the state, one can see the evidence of performance benefits to those attending smaller school settings. *Grade 6 data selected for purposes of mid-point performance between K-12.
| District/Indicator | Enrollment | G6 Math Proficiency | G6 ELA Proficiency | Graduation Rate |
| Buffalo | 30,674 | 9% | 29% | 76% |
| Rochester | 25,017 | 4% | 19% | 68% |
| Syracuse | 19,011 | 4% | 22% | 71% |
| Albany | 8,610 | 21% | 48% | 71% |
| Utica | 9,738 | 30% | 46% | 73% |
| Poughkeepsie | 4,154 | 2% | 38% | 60% |
| Elmira | 5,544 | 9% | 32% | 73% |
| Binghamton | 4,946 | 12% | 42% | 73% |
| Watertown | 3,972 | 31% | 58% | 77% |
| Niagara Falls | 6,572 | 12% | 32% | 64% |
| Amherst | 2,925 | 25% | 71% | 93% |
| Central Square | 3,580 | 11% | 35% | 87% |
| Scarsdale | 4,749 | 74% | 91% | 98% |
| Ballston Spa | 4,064 | 38% | 67% | 89% |
| West Irondequoit | 3,620 | 76% | 76% | 93% |
| Eden | 1,307 | 48% | 63% | 88% |
| Port Jefferson | 1,022 | 88% | 82% | 96% |
| Beekmantown | 1,915 | 30% | 60% | 94% |
| Cazenovia | 1,360 | 51% | 66% | 94% |
| Penn Yan | 1,312 | 49% | 54% | 87% |
Jones, Toma, and Zimmer (2006) found in a study of Texas high schools that larger districts were more likely to have a significant challenge in attendance rates of students, therein creating a challenge for student performance outcomes. This is referenced with regard to districts with greater geographical boundaries as well – that longer bus rides constitute risks for lower attendance rates (Cordes et al., 2022). When students are not in school, it is difficult to provide instruction and supports to meet academic standards and expected performance outcomes.
With regard to specific student performance on standardized test scores, Driscoll et al. (2001) found that students in California had lower standardized test scores if they were affiliated with larger districts. In particular, middle school performance appeared to be the most heavily impacted by a larger population. Driscoll also alludes to the idea that an indirect relationship between student performance and district size can be linked if there is a limitation in the local school decision-making authority, meaning that if a larger district uses the power of the central office to enforce programs or decisions that are consistent but lack the focus on targeted needs of an individual school building, this can be detrimental to student performance as well.
It seems straightforward or indisputable that, if student performance is negatively affected by district size, then the right path would be to alleviate the impact stemming from district size. But the size of districts across the country seems to elude this approach, with 25% of students being served by one of the largest 100 districts in the United States.
Ideal District Size – Community Involvement & Other Factors
While geography, money, and performance are the three largest factors that drive conversations around ideal school district size, there are other elements that are brought up in the discussion about whether larger or smaller district should prevail. One of these is the sentiments of community locus of control. In a 2013 study on school district consolidation, Boser listed that those against consolidation are concerned with the impact on the culture of their community (p.8). Having relationships with the staff in the school district is important to many families as their children traverse the K-12 educational journey. Further, there are real estate references brought into the conversation, that some people ascribe the value of properties to overall school performance. If larger districts perform lower than smaller districts, then property values will inherently be lower, which allows for stronger advocacy against consolidation and could easily be moved into the realm of arguing for smaller school district sizes. Even something as nuanced as the development of wellness policies have been identified as being more comprehensive in smaller school districts (Meendering, J. et. al, 2016).
How involved superintendents are in school improvement initiatives, overall school performance and programs, and their personal engagement with students and staff can also be related to district size. Hentschke et al. (2009) investigated several small city district superintendents and their behaviors in leadership, finding that those in smaller districts are more apt to be personally connected to elements of the school program, truly leading at the helm, a difference in behavior from that of larger districts, where the leadership of superintendents tends to be at such a high level it feels disconnected at times. For those community members seeking that personal touch with their school district, the leader at the top might be a critical piece. Smaller districts can offer connectivity in real time with a superintendent and their team, and the superintendent can find time in their schedules to be more “boots-on-the-ground” with initiatives.
Ideal District Size – Overall Recommendation
Overall, the success of students should drive the construct of ideal school size. Parameters that support the ideal school size that will help to promote school size should exist, such as boundaries for the maximum geographic size and the number of enrolled students and guidance to support what might be the best course of action when these maximums are exceeded.
It is clear that when all of the research on individual factors of school size are analyzed in a combined manner that smaller schools are more optimized for student success. If examined in isolation, some of these elements might make it seem as though larger districts can offer benefit to students, their families, school staff, and the surrounding community. But when examined in depth at a level that connects the elements together, it is clear that smaller districts offer a much greater array of services and supports to their student population that benefit the district and its students in the long run. Student performance is not the only factor, but it is certainly a large one if not the most important one, and when put next to the indirect correlations of geographical boundaries, long bus rides, and attendance as well as the minimal increased cost for the operations of a smaller district and the attraction of the culture that is created within a smaller school community, there should be no doubt that smaller is better.
References
101 American School Statistics: 2021/2022 data, Trends & Predictions. Research.com. (2022, October 17). Retrieved November 20, 2022, from https://research.com/education/american-school-statistics
Boser, U. (2013, August). Size Matters: A Look at School-District Consolidation. Center for American Progress. Retrieved November 21, 2022, from https://www.americanprogress.org/wp-content/uploads/2013/08/SchoolDistrictSize.pdf
Bureau, U. S. C. (2021, October 8). Census Bureau estimates show average one-way travel time to work rises to all-time high. Census.gov. Retrieved November 20, 2022, from https://www.census.gov/newsroom/press-releases/2021/one-way-travel-time-to-work-rises.html
Cavanagh, S. (2022, July 22). Personnel costs prove tough to contain. Education Week. Retrieved November 20, 2022, from https://www.edweek.org/teaching-learning/personnel-costs-prove-tough-to-contain/2011/01.
Cordes, S., Rick, C., & Schwartz, A. (2022). Do Long Bus Rides Drive Down Academic Outcomes? Educational Evaluation and Policy Analysis, 44(4), 689–716.
Driscoll, D., Halcoussis, D. & Svorny, S. (2001). School district size and student performance. Economics of Education Review, 22,
Hentschke, G. C., Nayfack, M. B. & Wohlstetter, P. (2009). Exploring superintendent leadership in smaller urban districts: Does district size influence superintendent behavior? Education and Urban Society, 41(3), 317-337.
Jones, J.T., Toma, E.F. & Zimmer, R.W. (2006). School attendance and district and school size. Economics pf Education Review, 27, 140-148.
Meendering, J., Kranz, E., Shafrath, T. & McCormack, L. (2016). Bigger ≠ better: The comprehensiveness and strength of school wellness policies varies by school district size. Journal of School Health, 86(9), 653-659.
Schiltz, F., & De Witte, K. (2017). Estimating scale economies and the optimal size of school districts: A flexible form approach. British Education Research Journal, 43(6), 1048–1067.
Schwartzman, D. (2017, June 27). When it comes to school districts, small is beautiful. California Policy Center. Retrieved November 20, 2022, from https://californiapolicycenter.org/when-it-comes-to-school-districts-small-is-beautiful/