The costs of online learning

Tamara Butler Battaglino, Matt Haldeman, and Eleanor Laurans

Online learning, in its many shapes and sizes, is
quickly becoming a regular part of the education experience for many of our
nation’s K-12 students. As it grows, educators and policymakers across the country
are beginning to ask the question: What does online learning cost?

A very
important question, to be sure, and one that our new Fordham-published paper seeks to address. Ultimately, new
technology-rich education models will need to be evaluated based on their
productivity—that is, the results that they produce relative to the required
investment. Unfortunately, within the nascent field of online learning, this information
simply isn’t yet available

This analysis is complicated—and needs to be—because
costs vary within digital education
just as they do within brick-and-mortar schooling options.

This analysis is complicated—and needs to be—because
costs vary within digital education
just as they do within brick-and-mortar schooling options. Educators and
policymakers pursue online learning for different reasons and adopt different
flavors of technology-rich models. Resource allocation also varies
significantly within these categories. Some models explicitly look for savings,
while others aim to free up resources from one area in order to use elsewhere.
Therefore we caution readers against looking for one simple “price tag” for
online learning or assuming that apparent savings necessarily translate into a
lower overall cost per pupil.

For schools that deliberately use technology to
reduce costs in one category so as to free up resources to invest elsewhere,
“savings” from online learning are often an important component of the
school’s overall resource-allocation strategy. We’re also mindful that economic
pressures could require all school models—both brick-and-mortar and online—to
cut costs (or keep them neutral) relative to today’s per-pupil-funding levels.
But we also need to recognize that a full range of needed experimentation with
online learning is going to require some investment and that some models worth
trying—and perhaps sustaining—may cost the same, or even more, than traditional
schooling.

A decade ago, the majority of online learning was
100 percent virtual in nature. Today, however, a wide range of online-learning
models can be seen. In our paper, we distill these into two broad categories:
virtual and blended. In virtual schools (examples include Florida Virtual, K12
Inc., Connections
Academy),
all instruction takes place online. Students, both full-time and part-time,
still interact with live teachers, listen to lectures, work on homework, ask
questions, and more, but all such activities occur at a distance, with
interactions facilitated by technology. In
blended schools, by comparison, students attend brick-and-mortar schools where
they alternate between online and in-person instruction. (Examples include
Carpe Diem, KIPP Empower LA, and Rocketship Education.)

The figure below outlines the variation in costs
of online-learning models versus the “traditional” model.[1] As is
evident from the data displayed there, the traditional-school model spends over
half of its budget on labor, with the majority of the rest put into school
operations. Content and technology costs combined are but a tiny fraction of
overall costs. A blended model, by comparison, has the potential to save approximately
$1,100 per student (11 percent). By significantly reducing the cost of school operations,
a virtual school can potentially save more: some $3,600 per student, a third of
the total cost of a traditional school. Bear in mind, however, that these
estimates reflect current variation in the field. They are no guarantee of equivalent
educational quality, given insufficient data on student outcomes associated
with the range of models.[2]

Estimated Per-Pupil Expenditures of Three
Schooling Models[3]

Schools’ resource-allocation strategies vary among
these models based on five primary cost-driving categories: labor, content
acquisition, technology and infrastructure, school operations, and student-support
services. Within each category, certain expenses are non-negotiable, while
others have the potential to leverage some cost savings.

1.  Labor

Labor (teachers and administrators), which
typically makes up well over half of district budgets, represents the largest
opportunity to rethink resource allocation and seek possible “savings.” Not
surprisingly, online schools vary widely in their human-capital structures,
including student-teacher ratios, salaries, and more. At its most basic level,
calculating labor costs per pupil is a simple equation with two variables: the
number of instructors and the average costs associated with those instructors (recognizing
that these may vary at different points in the school day). Reducing labor
costs via technology necessarily involves either decreasing the student-instructor
ratio or reducing average instructor compensation, typically by modifying the
overall instructor mix such as by combining the use of traditional teachers
with others such as paraprofessionals or aides.

2.  Content Acquisition and Development

Content costs in traditional brick and mortar
schools are relatively small compared to total per-pupil spending; some
districts spend less than $200 per student on instructional materials. In a
traditional school setting, content typically refers to the supporting
materials used by teachers for face-to-face instruction: textbooks, workbooks,
manipulatives, videos, instructional games, etc. Online schools, by comparison,
spend content dollars on tools that traditional schools seldom use, such as
data integration/management tools—which can be sophisticated and expensive. 

3.  Technology and Infrastructure

Technology, often a minimal portion of a
traditional school’s budget (about $200 per pupil), is, not surprisingly, a far
more significant expense for online schools over every genre.

4.  School Operations

From transportation to custodians to food
services, traditional schools spend significant sums (typically about 15 to 25
percent of total budgets) on non-instructional operations. In some cases,
virtual and blended schools have managed to reduce these costs to almost nothing;
in other examples, they actually pay more within these categories.

5.  Student-Support Services

Guidance counselors, special-education teachers,
and other student-support services (which typically cost up to about $800 per
student in a traditional setting) cannot be ignored when considering the costs
of online learning. At blended schools, such additional costs (mostly labor)
are more easily aligned with the regular school day. For virtual schools, these
support services can be significant expenses that require in-person visits (in
contrast to the main “virtual” program).

At day’s end, the promise of online learning is
twofold: more effective uses of technology have the potential both to improve
student outcomes and to create a more productive educational system. While our
paper works out the current costs of both virtual and blended models, it does
not tackle the question of productivity (i.e., how to improve and maximize
student achievement while keeping costs down). We hope it will encourage a focus
on better outcome data to help identify the most productive and effective
school models. Highlighting productivity is undoubtedly the first step towards
rewarding productivity and, over time, getting the most out of innovative
school models.

Tamara Butler Battaglino is partner of The Parthenon Group and co-founder and co-head of the Education Center for Excellence; Matt Haldeman and Eleanor Laurans are principal and senior principal of the Parthenon Group, respectively. They are the authors of the latest installment in Fordham's Creating Sound Policy for Digital Learning working paper series, "The Costs of Online Learning."


[1] The average per-pupil figure of
approximately $10,000 across the U.S., (not including central
administrative costs) combines all public-school types (elementary, middle, and
high school) without regard to district or state variations. The virtual-school
number represents an estimate for full-time high school students. The blended
figure represents middle school students, as some of the most promising blended
models are middle schools. Additionally, as scale can be a critical factor in
determining on-going costs, our cost numbers reflect schools with enrollments
of approximately 500 full-time equivalent (FTE) students.

[2] One important note: For online and
blended students, costs are sometimes reallocated in different ways that do not
mean savings from an overall systems standpoint. A virtual school, for
instance, may not charge for certain services (i.e., special-education
services) that are provided by the home school. This does not represent savings
from a system level; it simply means that costs are divided between the virtual
school and a student’s home school.

[3]
The cost figures outlined in the chart represent estimates gathered from
available public documents and conversations with experts and vendors within
the field. A wide variety of funding levels exist within online learning; for
example, research on virtual-school funding reveals that virtual schools are
operating on funding levels that range from below $4,000 per student to above
$9,000 per student. Given this fluctuation, we also provide an estimate for
variation within each figure (although there are outlier models at both the
high and low ends that do not fall within this band).