We set out to answer a basic (yet complicated) question: how much does each school in the D.C. metro area spend for each student it enrolls? What we found both surprised us and confirmed some of what we already knew. We expected to see that schools in different districts spend differently from one another, and indeed they do. As a group, Washington, D.C. Public Charter Schools (DCPC) have the highest per-pupil spending of the eight school systems in the area: $18,150 per pupil. Washington, D.C. Public Schools (DCPS), Alexandria City Public Schools, and Arlington County Public Schools all spend between $15,000 and $15,750 per pupil. Prince George's County spends the lowest: about $10,400 per pupil.

What's more surprising are the differences in spending within the same district. For instance, Tuckahoe Elementary in Arlington County spends about $11,310 per student, while Hoffman-Boston Elementary, also in Arlington, spends nearly twice as much: a little over $19,910 per pupil (see the School Districts section for more). We also found some patterns across types of schools that spend more or less than others (see the Similar Schools section for more). We encourage you to explore the site for additional information about school spending.

To learn more about why these differences may occur, we offer a primer in Drivers of School Spending. For an explanation of our accounting methods and data sources, see the Methods Summary.

Why do some schools spend more or less than others?

This is really two questions rolled into one. First, why do some districts spend more or less than others? Second, within a single district, why do schools spend differently from one another?

Differences in Public Revenue Drive Differences in Spending among Districts

Why is it that some school systems, such as the District of Columbia Public Schools (DCPS), spend more than others, such as the Prince George's County Public Schools? Simply put, districts that receive more money spend more money. Some districts have more money to spend for any number of reasons, and they spend what they get. Put another way, expenditures are limited by revenues.

Revenue for education comes from three public sources: local (typically via property taxes), state, and federal funds. If local property values are high and/or local tax rates are high, districts can raise more funds from local sources. If the state allocates more dollars to districts—or to districts with certain types of schools or students—then those districts receive more money from state sources. State allocations usually depend on a base amount per pupil, with extra funds for students based on their age, demographic characteristics, and/or special-needs status. Some states also give additional funds to districts with very low local revenue, which tend to be urban districts. If districts have more students eligible for federal aid (for example, Title I funds), they receive more federal dollars. Districts in the D.C. metro area are located in areas with different property values and tax rates, in different states, and with different demographics; hence, their revenues, and therefore their expenditures, vary. DCPS is an unusual case because it gets the majority of its funds from federal sources.

Although revenue can vary from year to year, mostly due to changes in the level of state funding, district leaders generally have a good idea of how much money they have in hand. Then they decide how to spend that revenue: how much to pay teachers and other staff, how many staff members to employ, what programs to offer, and so on. Although our analyses do not include revenue because it is accounted for and reported separately from expenditures, the U.S. Department of Commerce has revenue information for the nation's largest districts. Consider Fairfax and Montgomery Counties. According to the U.S. Department of Commerce, Fairfax County received less public revenue per pupil than Montgomery County in 2010–11, the most recent school year available. Not surprisingly, Fairfax County has a higher student-teacher ratio than Montgomery County (17.9 compared to 15.1), although the two have roughly equivalent starting teacher salaries. DCPS received the most revenue per pupil than any other district in the region; the district also offers teachers the highest starting salary in the metro area.

Individual schools also have the option of raising additional private revenue. For some, especially charter schools, philanthropic donations and/or sponsorship are critically important ways to generate funds for capital expenses. Other schools have particularly active parent booster clubs that fundraise for extracurricular activities, school beautification, or special programs.

Differences in Personnel Costs (Largely) Drive Differences in Spending among Districts and within Single Districts

What accounts for differences in spending among schools in the same district? Once district leaders receive their local, state, and federal revenue, they do not have to distribute those funds equally across schools on a per-student basis. Rather, personnel costs largely determine how much each school spends. Many personnel costs are the direct result of choices made by district leaders, such as setting salaries and maximum class sizes or requiring that each school have certain support staff. Some personnel costs are secondary effects of these choices. For example, districts determine yearly raises for teachers, and if one school attracts more experienced teachers, then it will spend more on salaries. A few, such as enrollment, are out of districts' control entirely.

Employee Base Salaries and Returns on Experience

Education is labor intensive, and personnel costs account for the largest chunk of school expenditures by far. Nationwide, about 80 percent of expenditures for public schools go toward employee salaries and benefits. Depending on the amount of revenue that districts have, this results in high or low salaries and, therefore, high or low school expenditures. Expenditures vary among districts as direct result of districts' employee salary rates. Consider that in DCPS, a first-year teacher with a bachelor's degree earned a starting salary of $51,539 in 2011–12. It's no coincidence that DCPS is among the higher-spending districts in the metro area. In Fairfax County, a comparable teacher earned about $7,100 less, and in Prince George's County, a comparable teacher earned $6,740 less. Although revenue is the primary determinant of staff salary levels, teacher salaries in particular can vary by district for a number of other reasons. Leaders of urban districts may choose to offer competitive salaries to offset the relative challenge of recruiting and retaining high-quality teachers. Another reason salaries in cities are high is because urban areas tend to have stronger labor unions, which are better able to negotiate with district leaders (or, some would argue, influence them) for higher pay.

Two schools in the same district that have identical numbers of students and staff should theoretically spend about the same, because salary is set at the district level. Yet some schools employ more seasoned teachers and others more novice teachers. Teacher-employment agreements have "salary scales" with built-in raises for each year of experience; most also have salary "bumps" for professional-education credits and advanced degrees. So while a first-year teacher with a bachelor's degree in DCPS earned $51,539 in 2011–12, a DCPS teacher with twenty-one or more years of experience and a PhD earned almost double that. Salary scales are the direct result of district decision making; because pay scales are set at the district level, individual schools have no control over base salaries and automatic yearly raises. (Maryland districts and DCPS determine base salaries via collective bargaining, and Virginia districts determine base salaries through district policies.) The experience level of teachers at an individual school and the personnel costs associated with more experienced teachers are the indirect result of that salary scale. Only if a district offers performance bonuses, as DCPS does, or ties yearly raises to satisfactory evaluations, does a school have any (small) say in how much to pay its teachers.

It is not just the experience level of the teacher that impacts spending: district rules that govern staff vacancies and, to some degree, teacher and principal discretion also play a part. Some district schools also have little control over which teachers get assigned to their buildings. Depending on the provisions of the collective-bargaining agreement, seniority rights may trump principal discretion relative to filling vacancies. If two schools have a vacancy in a fifth-grade classroom, one might be at liberty to hire a new teacher while the other may be assigned a veteran teacher at the high end of the salary scale. District rules can also state that if a reduction in force is necessary, less experienced teachers are the first to go. Both these scenarios result in a school having a more experienced workforce. More experienced teachers might prefer schools in a certain part of the district, or a particular principal might favor veteran teachers (or new ones) when filling vacancies. These decisions also affect the personnel costs at an individual site.

Charter schools are free to establish their own salaries, and there are no rules for filling vacancies. But because they must compete for talent, many pay their teachers at similar levels to nearby district schools.

True, a large part of salary levels are historical. Once a precedent for high or low salaries is set, it continues somewhat independently of the current district policies and leaders. In fact, when declining revenue limits teacher-salary levels and prevents district leaders from giving raises, leaders then turn to a second way to adjust personnel costs: staff-to-student ratios.

Student-Teacher and Student-Staff Ratios

Given the limitations of available revenue, one district might have higher salaries for teachers but opt to have larger classes, thus requiring fewer of them. Another might have lower salaries but smaller classes. One district with high revenue might have high salaries but larger classes, instead choosing to spend money on non-core classes such as art and music, or on programs for high-needs students. Finally, some districts might require (or be mandated by state law to ensure) that each school have one or more of a specific type of staff member—for example, librarians, nurses, counselors, special education coordinators, administrative assistants, and security guards. However, none of these decisions are made uniformly across the district, or else school-level expenditures would vary according to student population (for example, special education students) and be otherwise identical for all schools in a single district. Rather, uneven spending within a single district is a product of district leaders allocating staff positions to schools.

Sometimes the effects of staff allocations are obvious. For instance, in schools that serve primarily special education students, district leaders determine what types of teaching and medical support staff—and how many of them—are to be allocated to the building. Consequently, these schools have much higher per-pupil expenditures compared to traditional schools in the same district. The same can be said for alternative schools, which focus on students who have discipline problems, have been expelled from their home school, are pregnant or incarcerated, or exceed the age limit for high school. (See the filters for special education and alternative schools for more.)

Other times the effects of staff allocations are more nuanced. District leaders may choose to house certain programs and personnel at sites that meet specific criteria. For instance, district leaders might situate remedial academic activities and tutors at low-performing schools, place a specialized language or music program at a single site, offer high-cost electives such as engineering or manufacturing technology at a unique school, or assign more security guards at schools in higher-crime areas.

Finally, the size of the school affects the cost of staff-allocation rules. If a district requires one or more of a particular staff member per site, that rule is independent of the size of the school. Every school has a principal, for example, who would receive the same salary whether the school has 300 or 600 students. Every school also needs a nurse, a librarian, and other nonteaching staff; if a school is small, these costs can't be spread over as many students. While some districts are flexible—for example, sharing staff members across multiple sites—many are not. Large schools are also more responsive to change. Say a district sets a maximum class size in kindergarten of twenty students per teacher. If kindergarten enrollment declines by 15 percent at a school with only twenty kindergarteners and one kindergarten teacher, that teacher is still needed when the class size shrinks to seventeen. But if a larger school started with 160 kindergarteners and eight teachers, a 15 percent reduction in enrollment to 136 students means the school now only needs to staff and pay seven kindergarten teachers instead of eight.

View Metro Area Data

Region-Wide School-Level Accounting

The figures in the D.C. Metro Area School Spending Explorer represent school-level expenditures for the FY12 school year (2011–12), based on financial and enrollment data secured from local governments, school districts, and charter schools in the Washington, D.C. region. They include expenditures of both public funds (federal, state, and local monies) and private (philanthropic dollars, money raised by fundraising or booster club groups, and schools' reserves). The data are operating expenditures only and do not include capital expenditures and debt. The 2011–12 school year is the most recent for which we could get actual spending data when the analyses were begun. This is largely because expenditures are due at the end of the school year and must then be compiled and audited by the district, which results in significant lag time between when the money is spent and when the spending data are available.

To gather expenditure data, the Fordham Institute primarily used FY14 budget documents, which contain the audited actual expenditures for the 2011–12 school year. If a district did not report expenditures at the school level, we filed Freedom of Information Act requests for information on each district employee: title, salary, and the school where the employee worked. We used this information to calculate expenditures at each school. If personnel files provided the annualized salary instead of the actual salary paid, we found through district audits the totals for actual salaries paid, compared those numbers to the ones in the detailed employee files, and adjusted accordingly. In the event that a district's school-based expenditure documentation included salaries but not benefits for employees at the school, we first determined the fraction of school salaries to total salaries in the district. We then applied that fraction to the total paid in benefits for all district employees (as reported in the district budget). In this way we could assign a fractional portion of total district benefits to the employees at each school. Salaries not assigned to a school also received a portion of the benefit distribution, but are not counted as school-based expenditures.

For each district, we reviewed all expenditures not recorded at a school to determine if the expense directly benefited schools. These expenses included those related to employee benefits, transportation, building maintenance, utilities, and substitute and itinerant teachers.

Some functions—such as transportation, food service, and building maintenance—tend to be accounted for centrally, even though the schools receive the benefit of the service. Even districts that use site-based accounting have many expenditures that are not explicitly tied to specific schools, although those schools benefit. In the case of central expenditures, we allocated costs to the schools based on either the school's percentage of total district enrollment (for expenditures that directly benefited students), the school's percentage of a particular student population (for expenditures that benefited only a specific group, such as special education students, English language learners, or students receiving Title I services), or the school's square footage (for maintenance expenditures). Total and special-population enrollment data were compiled from a variety of state and district sources. Building square footage was obtained from district websites or by request. Not all central costs were allocated: for example, expenses related to operating the district central office are not included in the school-level expenditure totals because they do not directly benefit the schools.

Where possible, we combined programs at the same site that share teachers, administrators, and resources. For example, Woodson High School and Woodson Business and Finance Academy are combined into a single entry, as the latter is a school within a school and not a separate entity. Programs that do not operate as independent schools and are not associated with a host site are not included. (For example, we omit alternative programs that offer academic and social services but are not officially independent "schools" as classified by their districts.) Schools that share a facility but have separate administrators, teachers, and operating costs are kept as separate entries.

The data set contains all schools serving grades pre-kindergarten through 12 open during the 2011–12 school year, even if they have subsequently been closed. (Closed sites are noted.) The small number of schools for which we have insufficient data do not appear. Adult education programs are not included.

Schools are classified by grade level/type using grade-span information from the National Center for Education Statistics and cross-checked with district and school websites. For the map, the following designations are used:
• Early Childhood: pre-kindergarten only; pre-kindergarten and kindergarten only (even if the school serves primarily special education children)
• Elementary: highest grade offered is 6
• Elementary-Middle: lowest grade offered is 4 or earlier; highest grade offered is 8 or 9
• Middle: lowest grade offered is 5, 6, or 7; highest grade offered is 8 or 9
• Middle-High: lowest grade offered is 5, 6, or 7; highest grade offered is 9 or higher
• High: lowest grade offered is 9 or 10
• Special Education: schools that focus on special education students (these schools have no specific grade span and may serve only pre-kindergarten and kindergarten)
• Alternative: schools that only offer services to students who have discipline problems, have been expelled from their home school, are pregnant or incarcerated, or exceed the age limit for high school

All per-pupil expenditures at the district and school-type level (elementary, alternative, etc.) are totals, not averages. For example, to calculate per-pupil expenditures for Alexandria City Schools, we divided the total number of students (12,356) by the sum of all school-level expenditures ($185,878,796.01) for a district-wide per-pupil expenditure of $15,043.61. The same procedure was used to calculate the per-pupil expenditures of that district's twelve elementary schools. By presenting expenditures in this way, we do not give disproportionate weight to schools that have comparatively low enrollment. However, this method also obscures variation at the school level. For interested readers, the table below contains both calculations. The first column is what is used throughout this website. Large differences between the two calculations, such as those in DCPS, Montgomery County, and Prince George's County, are because those districts have a large number of high-spending, low-enrollment schools.

District	Total district expenditures divided by total district enrollment	Average of school-level per-pupil expenditures
Alexandria City	$15,044	$15,152
Arlington County	$15,599	$16,553
D.C. Public Schools	$15,743	$17,987
D.C. Public Charter Schools	$18,150	$18,173
Falls Church	$13,570	$13,436
Fairfax County and City	$11,704	$12,593
Montgomery County	$12,649	$14,118
Prince George's County	$10,408	$12,081
D.C. REGION (ALL)	$12,573	$14,424

Racial and ethnic demographic data and eligibility for the federal free or reduced-price lunch program are taken from the National Center for Education Statistics. In summary statistics, these data are averaged at the school level, as are special education percentages.

Income data are calculated using 2010 U.S. Census Bureau data on household income and geographic data on school catchment areas (obtained from each district's central office). To estimate household income for each school, we calculated the overlap between the school's catchment area and each census tract. Household income data were then weighted based on the percentage of the catchment area covered by each tract and then averaged. Because DCPS and DCPC use district-wide enrollment, this technique was not possible or needed there.

All seven districts and the D.C. Public Charter School Board were provided with an early view of the data and corresponding website and were given the opportunity to respond. Any errors and inconsistencies they noted in the data were fixed, and their feedback was integrated into the text where appropriate.

A Note on Calculating School-Level Expenditures

The question of "how much does a school spend?" is deceptively, and frustratingly, difficult to answer. First, because of the way that school districts are structured, school-level spending is hard to calculate. Technically, schools don't "spend" anything—districts do. The district pays teachers, the district purchases supplies, and the district provides transportation and food services. Principals have only a small amount of discretionary spending that is within their control, and that comes mostly from school-level fundraising rather than district funds. But unless a district uses site-based accounting, the compensation of the staff that works at a school full time is the only expense that is easily attributable to a single site. Beyond that, accounting is a bit muddier, to put it mildly. For example, buses are usually garaged and maintained at a central location. The mechanics that service the buses, the gas that fuels them, and the rent for the transportation depot are expenditures that aren't easily accounted for at the school level. As another example, districts may purchase instructional materials such as textbooks, large supplies like desks and chairs, and even operational needs from ceiling tiles to toilet paper in bulk, then distribute to schools as needed. These expenditures are clearly on behalf of students but aren't part of an individual school's budget.

Complicating the matter further is the fact that while school sites incur expenditures, districts may not count them that way. Of the districts in this database, only DCPS and the districts in Virginia use site-based accounting. Maryland districts do not. For Maryland districts (or any other districts that don't use site-based accounting), the first step to figuring out how much a school spends is to identify the experience level of the teachers who work there and how much they are paid. Other expenditures can be allocated out on the basis of a school's share of students. For example, if a district spends $1 million total on special education services, and if we know a particular school enrolls 10 percent of the district's special education students, then we can estimate that the school spends $100,000 on special education. However, districts are not required to report which students actually receive which services, how much those services cost, and where those students to go school. (Even Montgomery County's new site, created to "promote greater [budget] transparency," does not report operating expenses for the district as a whole, let alone for individual schools.)

Finally, districts are fundamentally complicated. Staff and students may move midyear, making calculating per-pupil expenditures difficult. One librarian or nurse might work at two or more schools, or a principal might oversee multiple, quasi-independent programs at a single location. Schools that enroll zero students can still incur expenditures—"closed" schools still require limited maintenance, or the district might be using the site for teacher training or as an office. Taken together, it was quite a task to determine how much each school spent. (Researchers and analysts such as Marguerite Roza, Karen Hawley Miles, and Bruce Baker have written on the herculean task of unraveling school finance.)

Alexandria City Public Schools

Alexandria City Public Schools conducts site-based financial accounting, so most of the expenditures were recorded in school-level budget documents. Centralized costs were allocated based on the students benefiting from the expenditure.

Enrollments (total, Title I, ESL, and special education) are from the Virginia Department of Education.

Arlington County Public Schools

Arlington County Public Schools conducts site-based financial accounting, so most of the expenditures were recorded in school-level budget documents. Centralized costs were allocated based on the students benefiting from the expenditure. Actual FY12 expenditure data were compiled from the Arlington Public Schools 2014 budget.

Enrollments (total, Title I, ESL, and special education) are from the Virginia Department of Education. Building square footage was compiled from a 2013 district bid document for pest-control services.

District of Columbia Public Schools (DCPS) and District of Columbia Public Charter Schools (DCPC)

DCPS figures were provided by the city budget office, which records site-based expenditures for all the district's schools. However, DCPS does not bear financial responsibility for retirement costs of its teachers. We used the retirement expenditures recorded by the U.S. Department of the Treasury and the D.C. Retirement Board to calculate and allocate expenditures on retirement benefits. We also identified and removed expenditures for retired employees so that they were not allocated to the schools.

Allocations of expenditures by the district that directly benefit the charter schools were gleaned from charter school financial data, which originates from audits. For comparability purposes, we excluded from the analysis the charter facility allowance, facility lease payments, and interest payments, because we do not include capital expenditures for district schools. Note that we do not report revenue, only expenditures. Charter schools that appear to outspend their district counterparts may be tapping into nonpublic dollars, such as reserves or monies earned via fundraising (which vary greatly across charter sites), because we know that charter schools receive less revenue than DCPS schools.

While we do not include the costs of operating the DCPS district office (among other central costs that do not directly benefit the schools), there are no such exclusions for DCPC because each school, or network of schools, operates independently.

Finally, more than half of the D.C. charter schools are in a network of two or more schools, meaning that they are operated together (almost as a mini-district). For these schools, only network-level expenditures are available. The per-pupil expenditure of each school in a charter network is calculated based on that school's share of the network's total enrollment.

Overall, the analyses for DCPS and DCPC should be considered the weakest in this study, as there may be centralized costs that could not be identified as specific to district or charter schools.

For both DCPS and DCPC, enrollments (total, Title I, ESL, and special education) are from the Office of the State Superintendent of Education annual enrollment audit. DCPS building square footage is from the city's Public Education Master Facilities Plan.

Fairfax County and Fairfax City School Districts

Fairfax County and Fairfax City have a joint agreement whereby the county system administers and provides services for the four Fairfax City schools. The Fairfax County Public Schools conducts site-based financial accounting, so most of the expenditures were recorded in school-level budget documents. Even some items posted as central allocations could easily be identified and allocated to a specific school because they were designated "SB" (for site-based). For other central expenditures, costs were allocated based on the students benefiting from the expenditure—special education expenditures, for example, were allocated based on the number of students receiving special-needs services at each school.

Enrollments (total, Title I, ESL, and special education) are from the Virginia Department of Education, and building square footage is from the Fairfax County Public Schools website.

Falls Church Public Schools

We used the FY14 budget documents for this analysis, which includes FY12 actual expenditures. Falls Church Public Schools conducts site-based financial accounting, so most of the expenditures were recorded in school-level budget documents. Even some items posted as central allocations could easily be identified and allocated to a specific school because they were designated "SB" (for site-based).

Enrollments (total, Title I, ESL, and special education) are from the Virginia Department of Education. Building square footage is from the district's Facilities Master Plan.

Montgomery County Public Schools

Montgomery County Public Schools does not generate site-based financial data for its schools. For this analysis, we relied on employee data provided by the district, which included annualized salary data for each employee at every work location in the district. We adjusted the annualized salary data based on actual employee costs recorded by the district and distributed benefits based on the percentage of each employee's salary compared to total salaries. We also used the 2014 budget document to evaluate what expenditures should be allocated to the schools, including some "centrally charged" employees who provide services to or work in multiple schools. These primarily consist of staff in food service, transportation, and building operations. When necessary, other costs were allocated based on the students benefiting from the expenditure.

Enrollments (total, Title I, ESL, and special education) are from the Maryland Department of Education. Building square footage is from the district's Facilities Master Plan.

Prince George's County Public Schools

Prince George's County does not generate site-based financial data for its schools. The analysis used employee data provided by the district, which included annualized salary data for each employee at every work location in the district. We adjusted the annualized salary data based on actual employee costs recorded by the district, and distributed benefits based on the percentage of each employee's salary compared to total salaries. The structure of payroll files indicates that some centralized salaries are recorded already at the school level. To avoid a duplicate count of salary expenditures, we did not allocate any centralized salaries unless we could determine that the salary file did not contain that same employee at an individual school. For other expenditures, we used material gathered from the Prince George's County Public Schools 2014 Budget, which includes 2012 actual expenditures. In some cases, we relied on 2012 budget detail for allocations when the 2014 budget document provided insufficient detail on 2012 expenditures.

Enrollments (total, Title I, ESL, and special education) are from the Maryland Department of Education. Building square footage for schools was compiled from the FY12 Comprehensive Annual Financial Report; for non-school sites, this information was provided by Prince George's County Public Schools.

We were unable to report expenditures of the eight charter schools in Prince George's County. The direct costs listed in the Prince George's County Public Schools 2014 Budget include some expenditures by charter schools. However, these figures do not capture 100 percent of the expenditures at any of the charter schools. We requested the 2012 audits for the charters for a more accurate portrayal of their expenditures. However, at the time we completed our analyses, those audits had not been completed.

All website content and analyses are products of the Thomas B. Fordham Institute.

Accounting: Larry Maloney, Aspire Consulting
Website Text and Analysis: Dara Zeehandelaar
Website Design: Victor Hernandez, Joe Portnoy and Matt Richmond (Fordham)

If you encounter any errors or wish to contact the authors or request raw data, please email [email protected].

Metro Region (All Districts)

Alexandria City Public Schools

Alexandria City Public Schools is a small but diverse district located across the Potomac River from southern Washington, D.C. The district has nineteen schools, including T.C. Williams, its only high school. In total, district schools spend $15,044 per pupil, the fourth highest in the D.C. metro area after D.C. charter schools, DCPS, and Arlington County. Alexandria's big spender is Jefferson-Houston, a school serving grades pre-K–8; although the school is one of the smallest in the district, enrolling only 366 students, it spends $20,826 for each of them. The district reports that a high special education population there (27 percent of enrolled students, according to state data) is the primary driver of expenditures, and despite its small size the school employs fifty-eight licensed staff members. The district's most frugal school was also one of its largest, with one of the lowest percentages of special education students (7 percent): William Ramsay Elementary spent $12,734 for each of its 743 students.

Arlington Public Schools

The thirty-two schools that make up Arlington Public Schools spend $15,599 per pupil. This makes it the highest-spending locality in our analysis after DCPS and D.C. charter schools. Arlington has similar demographics to nearby Fairfax County, but spends nearly $4,000 more per pupil than Fairfax (which is especially interesting because Fairfax schools are, on average, located in areas with higher income). Arlington has a significantly higher percentage of white students and lower percentage of students eligible for the federal free or reduced-price lunch program than two of its nearest neighbors, DCPS and Alexandria City, but spends nearly the same amount per pupil as those two districts.

Tuckahoe Elementary, Nottingham Elementary, and Taylor Elementary are Arlington's thriftiest, each spending less than $12,000 per pupil. The three are also among the largest elementary schools in Arlington. Among its traditional schools, the district's biggest spenders are Hoffman-Boston Elementary ($19,913 per pupil), Randolph Elementary ($19,675), and Kenmore Middle ($19,566). The site with the highest expenditures overall is the Stratford Program, a secondary school for students with special needs ($45,663 for each of its forty-six pupils).

District of Columbia Public Charter Schools

Charter schools in Washington, D.C. spend $18,150 per pupil—more than any of the seven districts in the D.C. metro region. Compared to DCPS, D.C. charters serve significantly more students eligible for the federal lunch program than does DCPS (with a school-level average of 74 percent compared to 56 percent), but a smaller percentage of special education students (12 percent compared to 17 percent). Charters are mostly located in the less-affluent regions of the city.

Only one high-spending charter school, St. Coletta, specifically serves students with severe special education needs ($94,565 for each of 234 students). Other big spenders include Options Public Charter School, which spends $51,910 per pupil; the school offers programs for high-needs special education students and at-risk students, but also serves general education students. Another, the SEED school ($41,795), is a full-time residential program. The lowest-spending charter schools include Booker T. Washington, a now-closed school that focused on job training for high school students and adults ($12,337); Excel Academy, D.C.'s first charter school ($13,145); and Washington Latin Middle and Upper Schools, which are among the highest-performing in the city ($13,183).

About two-thirds of D.C. charter schools are in networks of two or more schools, meaning they are operated together (almost as a mini-district). Per-pupil expenditures of charter networks are generally comparable to stand-alone charters. The Roots and Washington Latin networks are relatively thrifty; the E. L. Haynes ($25,051) and Maya Angelou ($27,448) schools have some of the highest expenditures in the area. Like all other schools, charter network expenditures are largely related to the programs they offer and the students they serve. For example, one school in the Maya Angelou network is an alternative school that offers at-risk students wraparound services, such as counseling and job training, in addition to academics.

For comparison purposes, all expenditures in the region are operating expenditures only. Charter schools in D.C. are notoriously underfunded when it comes to facilities, and the school-level per-pupil expenditures shown here exclude capital outlay. However, for comparison's sake, we can also divide total expenditures by DCPS and D.C. charters by the number of students in each sector. Total expenditures include capital outlay, central office expenses, adult education, debt servicing, and all other costs not directly associated with individual schools. Doing so yields $32,353 per-pupil expenditures in DCPS, and $20,715 for D.C. charters. (On capital and debt alone, DCPS spent $6,294 per pupil.)

Note: For charter networks, we have total expenditures for the entire network, but not for the individual schools in that network. The per-pupil expenditure of each school in a charter network is calculated based on that school's share of the network's total enrollment.

District of Columbia Public Schools

D.C. Public Schools is widely known to spend (and receive) more per pupil than most other districts in the nation; with a total of $15,743, DCPS is in the middle of the regional pack. However, the average per-pupil expenditure is $17,987, a number that is driven up by large number of high-spending, low-enrollment schools in the district. (See "Region-Wide School-Level Accounting" in the Methods Summary, where we discuss the different calculations.) It's not surprising that the big spenders are among the district's special education centers, including Mamie D. Lee ($58,564 per pupil), the now-closed Prospect Learning Center ($63,182), and Sharpe Health School ($72,906), all for students with severe special needs. (Note that the Transition Academy at Ballou, a special education program for students with emotional disabilities, is not included because many of its expenditures cannot be separated from Ballou High School.)

C.H.O.I.C.E. Academy, an alternative school serving students in grades 6–12 who have been expelled from their neighborhood schools due to discipline problems, has the highest per-pupil spending in the district at $113,013 for each of its twenty-three students. However, two other alternative schools, the Washington Metropolitan High School and Luke C. Moore high school, have relatively low per-pupil expenditures ($13,538 and $14,955, respectively).

Of high-spending traditional schools, Eastern High School enrolled only 303 students in 2012; it was in the process of reopening on a grade-by-grade basis, so the fact that it was under-enrolled drove up costs. The traditional schools that spend the least per pupil include Deal Junior High School ($10,792), and Murch, Janney, Mann, and Watkins elementary schools (all less than $12,000). These schools have percentages of special education students and students eligible for the federal free or reduced-price lunch program that are well under the district average. They also have higher-than-average enrollments.

DCPS does have a reputation for high teacher salaries, which helps explain its per-pupil expenditures. But compared to other districts in the region, DCPS's student-teacher ratio is in the middle. So while it might have high teacher salaries, it has fewer teachers than some of its neighbors. DCPS also has concerns beyond teachers: according to the district's website, "we must invest a significant portion of our budget on required supports such as food services and school security."

Fairfax County and Fairfax City School Districts

Fairfax County and Fairfax City have a joint agreement whereby the county system administers and provides services for the four Fairfax City schools. Combined, Fairfax has the largest enrollment of any district in the Washington, D.C. metro area, and thirteenth-largest enrollment in the nation. Montgomery and Prince George's counties have about the same number of schools, but fewer students. The district is also one of the region's thriftiest, spending $11,704 per pupil—only Prince George's County spends less. Among the schools that serve a traditional population, the three that spend the least are Louise Archer ($8,980), Haycock ($9,146), and Greenbriar West ($9,392) elementary schools. (Dunn Loring, Rocky Run, and Virginia Hills Elementary Resource Centers spend less, each about $8,250 per student, but they are early childhood education centers for children with special needs.)

The district's highest-spending traditional schools are Armstrong ($14,449), Lemon Road ($15,426), and Bucknell ($17,548) elementary schools. The high-spending schools are among the district's smallest, while the low spenders have around 800 students each (compared to an average elementary-school size of 678, with twenty-three schools at 850 students or more), which suggests that large schools in Fairfax are more efficient, but only up to a certain point.

In the region, Fairfax schools are the most uniform across grade levels, with virtually no difference among elementary, middle, and high schools. Compared to other districts, Fairfax also spends less in its seven schools that serve students with special needs—a total of $32,466 per pupil is less than similar schools in all the other metro area districts save Prince George's County (and comparable to DCPS). That figure doesn't include the district's three early childhood centers, which also focus on specialized education for children with developmental delays. Those three sites spend the least of all early childhood centers in the region, despite serving a special population of students.

Falls Church City Public Schools

Falls Church is the region's tiniest district, nestled between Arlington and Fairfax Counties. The district has only four schools: Mount Daniel (early elementary), Thomas Jefferson (late elementary), Mary Ellen Henderson (middle), and George Mason (middle-high). The district spends $13,570 per pupil. Interestingly, only Thomas Jefferson is located within the city boundaries itself; the other schools are physically located in Fairfax County. Falls Church also enrolls a small number of tuition-paying students who come from outside the city boundaries, meaning that students can pay to attend if they do not live within district. These students, whose families pay rates comparable to the district's per-pupil expenditures on its own students, are also charged for the actual costs of special education and/or gifted-and-talented services.

Montgomery County Public Schools

Montgomery County Public Schools has the largest enrollment of any district in Maryland, and among the largest in the D.C. metro area. It maintains approximately the same number of schools as Fairfax and Prince George's counties, and at $12,649 per pupil spends slightly more than those two districts. Like many of its neighbors, Montgomery County middle schools spend slightly more than the elementary and high schools, and its alternative and special education programs spend most.

Not surprisingly, the big spenders in Montgomery County are the six campuses that serve students with special needs. As a group, they spend $48,225 for each of their 494 students. The county also has one alternative site, the Blair Ewing Center, that hosts six programs for middle and high school students; it spends $35,812 for each of its 185 pupils.

Ten Montgomery County schools spend less than $10,000 per student; all of them are elementary schools. Most of the ten are among the largest elementary schools in the district, with enrollments well over the 534-student elementary school average. The low spenders are also mostly located in high-income areas, and serve a comparatively small percentage of special education students and students eligible for the free or reduced-price lunch program. These campuses include College Gardens ($8,897), Bradley Hills ($9,067), and Wood Acres ($9,311). For comparison, there are twenty-two elementary schools with 400 or fewer students; as a group, they spend an average of $14,162. That group includes Washington Grove Elementary, one of two traditional district schools that top $17,000 per student (the other is Sligo Middle).

Prince George's County Public Schools

Prince George's County has more schools than any of the other districts in the D.C. metro area, and spends the fewest dollars per pupil ($10,408). The discrepancy in dollars is even more striking when compared to its nearest neighbor, DCPS (which spends $15,743 per pupil), considering that the two districts have similar demographics. One major difference is scale. The 191 district schools in Prince George's County have an average enrollment of 628 students, while the 120 DCPS schools enroll an average of 365 pupils. Another key difference is revenue. Prince George's County simply receives less money per pupil, so it creates economies by having larger schools and one of the region's highest student-teacher ratios.

As with the other D.C.-area districts, the Prince George's County schools with the highest per-pupil spending are among the seven that enroll students with special needs: three early childhood centers and four in the higher grades. These sites include the Tanglewood ($65,987), C. Elizabeth Rieg ($47,162), and Margaret Brent ($37,736) Regional Centers. The four alternative schools, which serve students who have discipline or other problems or who would benefit from a non-traditional education, have high expenditures as well; Croom Vocational ($28,212), Green Valley ($28,529), and Annapolis Road ($36,863) Academies are examples. Of the traditional district schools, those with the lowest expenditures in the district—and the entire region—include Barack Obama, Vansville, and District Heights elementary schools (all under $8,000).

Big Spenders

The highest-spending schools in the region are a diverse group. (Click on the link below to see a list of the lowest-spending primary and secondary schools in the D.C. metro area, as well as the three lowest-spending schools in each district.)

The highest-spending primary schools in the D.C. metro area are spread across Washington, D.C., (both district and charter schools), Prince George's County, Alexandria, and Arlington. These schools appear to have little in common. Some serve a much higher percentage of students eligible for the federal free or reduced-price lunch program than others (although all have approximately 30 percent or more of their students eligible). Some schools are more diverse, others are not. Yet they do have a few characteristics that link them together. All but one (Judith P. Hoyer Montessori) are eligible for federal Title I funds. Many have a very high percentage of special education students, with most close to 20 percent or higher. And they are all relatively small—each is below the metro-area average of 587 students per primary school. (The E. L. Haynes network consists of two campuses.) This suggests that there's a "sweet spot" for enrollment, and schools that are too small are at a disadvantage relative to efficient use of personnel and resources. It also suggests that schools with higher-needs students are receiving additional funds.

All of the region's highest-spending secondary schools are located in Washington, D.C.: five district schools and six charter schools (the Maya Angelou network has two separate campuses). All are eligible for Title I funds except Spingarn Senior High School, and have student populations that are nearly completely African American, or African American and Hispanic (Cardozo High School and Shaw Middle School). They share three other characteristics. As a group, they enroll high percentages of students eligible for the federal free or reduced-price lunch program. Like the high-spending primary schools, most have very high percentages of special education students (the exceptions being SEED PCS, which is a residential school, and Thurgood Marshall Academy PCS). And like the high-spending primary schools, they are small. Each is significantly smaller than the metro-area average of 1,160 students per secondary school.

The three highest-spending traditional schools in each district are a diverse group. They include elementary, middle, and high schools, as well as mixed grade spans. Thirteen of the twenty-one are Title I eligible; eight are not. Three are magnet schools (Cora Kelly in Alexandria, and the two Montessori schools in Prince George's County). Most are located in lower-income areas but some are not. Unsurprisingly, their common characteristics are size and special education enrollment. All of these schools are well below the average enrollment for schools of similar grade span in each district. For example, the average enrollment for elementary schools in Fairfax County is 678 students; the three highest-spending schools in that district—all elementary schools—have far smaller enrollments. The highest-spending schools in each district also have comparatively higher percentages of special education students than other schools in their same district.

Lean Schools

There are few surprises about which schools make the list of lowest-spending schools in the region. (Click on the link below to see a list of the lowest-spending primary and secondary schools in the D.C. metro area, as well as the three lowest-spending schools in each district.)

The lowest-spending primary schools in the D.C. metro area are all in Prince George's County, which as a district spends the least per pupil. The low spenders are all fairly large, with enrollments near or above the metro-area primary-level average of 587 students (suggesting that an efficient school should be large enough to enjoy economies of scale for its operations, but not so large as to become unmanageable). None of the schools on this list receive Title I funds, which also at least partially explains their low expenditures. Simply put, they do not receive Title I funds, so they do not have that money to spend. In the low-spending primary schools, special education enrollment ranges from 6 to 10 percent, which is well below the district (and region) average. Otherwise they are a diverse group, both in terms of student demographics and percentage of students eligible for the federal free or reduced-price lunch program.

All ten of the lowest-spending secondary schools in the region are also in Prince George's County. Like the low-spending primary schools, none receives Title I funds. The schools are otherwise a diverse group: some have a high percentage of Hispanic students (High Point and Bladensburg) and others have a relatively high percentage of white students (Eleanor Roosevelt, Bowie, and Samuel Ogle). Eight of the ten schools have a percentage of students eligible for the federal free or reduced-price lunch program lower than the district average, and all have a lower percentage of special education students. Like the low-spending primary schools, this group of secondary schools shares another feature: size. Most are near or exceed the area-wide average of 1,160 students per secondary school.

The three lowest-spending traditional schools in each district have one major similarity: they are almost all elementary schools. Most of them also enroll a far smaller percentage of students in the federal lunch program than the average in their district. Two-thirds of the schools in this low-spending group have less than 30 percent of their students enrolled in the program, compared to a metro-area average of 47 percent. Almost none of them are Title I eligible (William Ramsay in Alexandria, Booker T. Washington PCS, and Excel Academy PCS are the exceptions). And they are all moderately sized, with enrollments at or near the region-wide average for traditional schools.

Early Childhood Centers

In the D.C. region, Fairfax County, Prince George's County, DCPS, and D.C. charter schools have campuses specifically dedicated to early childhood education. (The other districts also offer pre-kindergarten, although not at unique sites. In these cases, we could not disaggregate spending on early childhood education from expenditures on the other elementary grades at the school. We therefore report expenditures on early childhood education as part of overall per-pupil expenditures at that school.) Expenditures by these centers vary throughout the area.

Early childhood centers in Fairfax and Prince George's Counties serve primarily or solely special education students. This is reflected in the expenditures in Prince George's County: the three early childhood centers spend between $30,000 and $40,000 per pupil, which is significantly more than the district spends in its traditional schools. For example, Chapel Forge in Prince George's County spends $39,239 per student; the center focuses on children showing signs of developmental delays or who need specialized care and therefore employs a team of therapists and aides as well as teachers. Compare this to the three early childhood centers in Fairfax County, which enroll only children with special needs: each spend approximately $8,200 per student. It is not surprising that the two districts are located in different states: the requirements of preschool programs are different across states and vary by funding source, so early childhood centers in Maryland districts may be required to (or may choose to) offer different services than those in Virginia.

Bridges PCS was likely a high-spending school during the 2011–12 school year because at the time the school was only enrolling preschool students (and not operating at capacity). It has since expanded to include other elementary grades.

Note that these analyses include only those programs that are operated by school districts.

Special Education Schools

Most districts in the D.C. metro region have schools that only serve its highest-need special education students. These students have needs that cannot be met at a general education setting at a traditional school. They may, for example, require significant instructional adaptations, assistance performing basic tasks, one-on-one clinical or medical care, and/or specialized facilities. Student diagnoses can range from autism to vision or hearing impairments, from language disorders to physical disabilities, and from behavior disorders to health impairments. Because of the special needs of the students they serve, these schools generally have low enrollment and a large number of staff, which means their per-pupil spending is high. Schools may also serve students who reside in neighboring districts. St. Coletta, for example—a Washington, D.C., charter school that spends $94,565 per pupil—accepts students from Washington, D.C., Maryland, and Virginia.

Sites that are identified as special education schools are categorized that way on district websites, and in some cases school staff confirmed the primary population served.

Alternative and Vocational Schools

Some districts in the D.C. area have schools that are designated as "alternative." They serve students who have discipline problems, have been expelled from their home school, are pregnant or incarcerated, or need non-traditional options. These schools run the gamut in the resources and services they offer, the number of staff that provide those services, and how much those resources cost per pupil. The C.H.O.I.C.E. Academy, a DCPS school that serves only twenty-three students with long-term suspensions or expulsions, has expenditures of $113,013 per pupil. This school appears to be an outlier, however—most of the other DCPS alternative schools spend similarly to the district's traditional schools. The four alternative schools in Prince George's County, on the other hand, spend far more per pupil than traditional schools in that district.

Note that adult schools are not included in this analysis.