Public education is changing rapidly, and K-12 finance systems must adapt to a new era with fewer students and more competition. Since the start of the COVID-19 pandemic in 2020, public schools have lost 1.4 million students and are projected to lose another 2.7 million by the 2031-32 school year. And with the rise of school choice policies such as education savings accounts, charter schooling, and open enrollment, a family’s education options are no longer determined solely by where they can afford to live.
States’ school finance systems, many of which date back decades, must be modernized. Central to this is how they count students for funding calculations, a key policy detail that in many states, including those with school choice programs such as Arkansas, Tennessee, and West Virginia, strains state budgets and provides the wrong incentives.
Three best practices can guide policymakers in designing a sound approach to counting students. The following analysis details these three best practices, offers models for policymakers to consider, and compares how 15 states count students.
Best Practice #1: Use Current-Year Enrollment Counts
The most important policy feature in funding calculations is the timeframe used to count students. The three approaches used by states, along with their tradeoffs, are described below.
- Current-Year Enrollment: States such as Arizona, Indiana, and Texas use current-year enrollment counts to calculate funding. This approach, which counts students attending public schools in the current year, is most responsive to enrollment changes but provides the least budget predictability for school district administrators because districts’ actual funding isn’t known before the school year.
- Prior-Year Enrollment: States such as Alabama, Iowa, and New Hampshire use prior-year enrollment counts to calculate funding. This approach, which gives school districts funding for the number of students they served the year prior, provides greater budget stability but is less responsive to enrollment changes than using current-year counts. Districts that lose students from one year to the next receive funding for “ghost” students (i.e., students who are no longer enrolled, but for whom districts still collect funding), while districts that gain students don’t get funding for these additional students they’re serving until the following year.
- Declining Enrollment: States such as Kansas, Missouri, and Montana use the highest enrollment counts of two or more school years, usually including the current school year. This hold-harmless approach gives school districts the longest off-ramp for adjusting budgets to enrollment changes, but it usually results in the most ghost students, making it more expensive than using current-year or prior-year enrollment counts.
Current-year enrollment counts are the best policy for modern K-12 education systems. Under this approach, funding is based on the students currently in classrooms, and dollars follow students without delay to school districts or charters with growing enrollment.
The rationale for using current-year enrollment is straightforward: state funding should reflect school districts’ costs, which are driven by student enrollment. Funding “ghost students” diverts dollars from school districts with stable or growing student populations and diminishes the incentive for declining-enrollment districts to adjust budgets or respond to students’ needs. And in an era defined by increasing school choice policies, current-year enrollment also avoids double-funding choice participants.
It’s true that some school district costs are difficult to adjust in the short term. For example, a school district that loses 25 students mid-year across different schools and grade levels can’t reduce its teaching staff to trim expenses. Using prior-year or multi-year enrollment counts helps smooth revenue fluctuations, giving administrators more time to adjust costs. But these policies are costly, and their benefits are overstated.
For example, a study published by Reason Foundation estimates that Missouri’s declining-enrollment provision, which allows school districts to use the greatest of the current year, prior year, or second prior year student count, cost the state an additional $197 million in 2022, or 4.7% of the state’s formula funding that year for students no longer there.
The returns on this investment are questionable. Many school districts, and every private school in the country, are funded based on students currently in classrooms and manage budgets without issue. And California, which has a generous declining-enrollment provision that cost an estimated $4 billion for ghost students in 2023, funds charter schools using current-year counts.
While some staffing costs are fixed in the short term, school districts can adapt to enrollment changes using tools such as forecasting, budget reserves, and, if necessary, trimming non-instructional expenses. At a time when more than half of all public school staff are non-teachers, public school costs aren’t nearly as fixed in the short term as sometimes claimed.
Best Practice #2: Use Average Daily Membership to Count Students
Policymakers must also decide how to count students. The three approaches to counting students, along with their tradeoffs, are described below.
- Average Daily Membership: States such as Arizona, Minnesota, and Tennessee use average daily membership (ADM), which measures average student enrollment over a given period. ADM is calculated by aggregating the number of students enrolled each day and dividing that total by the number of instructional days during that period. ADM is by far the most accurate approach, tying education dollars directly to students and capturing daily enrollment. Its main drawback is that it’s more complex to administer than using simple count dates.
- Average Daily Attendance: Average daily attendance (ADA), which is used by states such as Texas, Idaho, and Missouri, is like ADM but with one key difference: only students who attend school on a given day are counted, creating a direct link between funding and attendance. ADA incentivizes attendance, which research links to student outcomes. However, ADA is less transparent, can complicate school choice funding, and there’s little research showing that ADA itself improves student outcomes. ADA also puts school districts and charter schools that serve low-income students at a fiscal disadvantage, since these students have worse attendance on average.
- Count Dates: States such as Indiana, Iowa, and West Virginia use one or more count dates to capture student enrollment. The count-dates approach is transparent and easy to administer, but it fails to capture enrollment changes that occur between count dates and after the final count date. For example, a school district that enrolls a student who switches from a charter school after the last count date would not receive funding for that student (and the charter school would not lose funding) until the following school year.
Average daily membership’s primary benefit, accuracy, outweighs its administrative complexity, making it the best option for a modern funding system. While the simplicity of using count dates is appealing, and this approach is superior to average daily attendance, it is less sensitive to enrollment changes than ADM, a notable disadvantage in an era where students increasingly change schools.
Best Practice #3: Account for the Entire School Year
Finally, policymakers must consider the timeframe to count students within a school year. This feature varies most across states, even among those that count students in the same way.
For example, Texas and Idaho both use current-year average daily attendance, but Texas calculates enrollment over the entire school year, while Idaho uses only the highest 28 weeks. Similarly, Iowa and Montana both use count dates, but Iowa uses one snapshot (October) while Montana uses an average of two snapshots (October and February).
States should capture enrollment for the entire school year, not just a portion of it. This ensures that funding is responsive to all enrollment changes, making each day count and rewarding districts and charters that attract and retain students throughout the school year.
State Models for Policymakers to Consider
For state policymakers, the most critical policy feature is to adopt current-year enrollment counts, as Arizona and Indiana have done. These and other states show that school districts can adapt to enrollment changes in real time without hold-harmless funding, including declining-enrollment counts.
- Arizona uses current-year ADM to allocate K-12 funding. However, it considers only the first 100 school days, falling short of tracking students for the full school year. The Arizona Department of Education distributes aid monthly, using the latest available data and entitlement calculations. The first two payments (July and August) are based on prior-year ADM, and monthly payments are the difference between a rolling share of the annual aid and year-to-date payments. Arizona completes a final statewide calculation in July and adjusts the following school year’s aid payments, as needed.
- Indiana uses current-year data with two enrollment count dates: an October count that determines funding for July to December, and a February count that determines funding for January to June. As such, enrollment changes after the February count don’t affect funding levels. The Indiana Department of Education (IDOE) distributes at least 12 payments annually, according to a payment schedule set each year by the State Budget Agency and the Governor’s office. IDOE uses estimated membership data until actual figures are available, with periodic adjustments to reconcile any discrepancies. If calculated state aid distributions exceed appropriated K-12 funding, the state can transfer up to $25 million from a reserve account to cover the difference.
While many states already use ADM funding, most base funding on prior-year or declining-enrollment counts.
Table 1 compares 15 states with Reason Foundation’s three best practices. Note that some policy nuances are excluded for clarity and, while some hold-harmless provisions are identified in this table, it is not intended to be a full account of all such policies across and within states.
Table 1: Comparing Approaches to Counting Students in 15 States
State | Description | Current-Year Enrollment | Average Daily Membership | Entire School Year Data |
Alabama | Alabama determines school districts’ average daily membership based on the average number of students enrolled during the first 20 days after Labor Day in the prior year. | No | Yes | No |
Arkansas | Arkansas uses prior-year average daily membership, but school districts with an enrollment decline over the two preceding school years receive declining-enrollment funding. Arkansas calculates average daily membership using counts from the first three quarters of each school year. | No | Yes | No |
Florida | Florida uses current-year enrollment with two primary survey periods, October and February. Florida also has a declining-enrollment stabilization fund that, among other things, partially covers the funding gap when actual enrollment falls below state budget projections. | Yes | No | No |
Idaho | Idaho uses current-year average daily attendance, basing funding on the attendance period ending on the first Friday in November (Salary and Benefits Apportionment) and the highest 28 weeks of the school year (discretionary funds). Idaho also has a declining-enrollment provision that allows districts with enrollment declines to use their prior-year average daily attendance, less 3%. The cost of this provision is spread across school districts. | Yes | No | No |
Iowa | Iowa uses prior-year enrollment counts on October 1st, and provides hold-harmless funding equal to the difference between budget-year regular program funding and 101% of its prior-year regular program funding. | No | No | No |
Kansas | Kansas law allows school districts to use the greater of the current-year enrollment count on September 20th or the prior-year enrollment count on the same date. | No | No | No |
Missouri | Missouri law allows school districts to use the greater of the weighted average daily attendance for the current year, the prior year, or the second prior year. Missouri also has two additional hold-harmless provisions: districts with a prior-year ADA greater than 350 are guaranteed at least their per-pupil state aid amount in 2005-06, while those with a prior-year ADA of 350 or less are guaranteed at least their state aid amount in the higher of either 2004-05 or 2005-06. | No | No | Yes |
Montana | Montana uses the higher of prior-year enrollment or a 3-year average. Enrollment is determined by averaging two count dates (October and February). | No | No | No |
New Hampshire | New Hampshire uses prior-year average daily membership, basing counts on the entire school year. | No | Yes | Yes |
Oklahoma | Oklahoma law allows school districts to use the greater of the weighted average daily membership of the first nine weeks of the current school year or the prior school year. | No | Yes | No |
Ohio | Ohio law allows school districts to use the greater of the average daily membership of the prior year or the average of the prior three years. | No | Yes | Yes |
Tennessee | Tennessee uses prior-year average daily membership, averaged across nine reporting periods. | No | Yes | Yes |
Texas | Texas uses current-year average daily attendance, based on the entire school year. Texas has a hold-harmless provision for districts that lose at least 2% of their ADA, providing funding up to 98% of prior-year ADA. This funding is subject to appropriation. | Yes | No | Yes |
Utah | Utah uses prior-year average daily membership plus a growth factor. The growth factor is based on current-year ADM on October 1st, giving credit to school districts with growth and holding harmless districts with declining enrollment. Utah also has a separate hold-harmless provision, subject to available funding, for districts where ADM drops more than 4% below the average for the highest two of the preceding three years. | No | Yes | Yes |
West Virginia | West Virginia uses prior-year enrollment in the second month of the school year. | No | No | No |
Conclusion
Public education is changing rapidly, and states must reform how they count students for funding purposes. Declining-enrollment policies, a form of hold-harmless, create ghost students, are unnecessary, and are a poor use of finite education dollars.
In a new era of K-12 finance focused on better serving students, adopting a policy that combines current-year enrollment counts, average daily membership, and full-year accounting will provide strong incentives, ensure compatibility with choice programs, and save states money.