Is ESA draining ICCSD enrollment?

Education Savings Account use vs. actual movement from public to private · ICCSD boundary (Iowa DE district 3141), 2022-23 to 2025-26

A high voucher count is not the same as students leaving. ICCSD has 1,440 resident students on an Iowa Education Savings Account (ESA) voucher in 2025-26. That number only means a public loss if those students would have enrolled in ICCSD without the program. Most of them would not have. They were already in private school.

ICCSD-resident ESA users (2025-26)

1,440

headline voucher count

Already private

~78%

about 1,130 students, no enrollment effect

Real move from public to private

120-190

over 3 years, under 1.5% of the district

From 1,440 vouchers down to about 175 transfers

The headline ESA count cannot show a transfer by itself. Most of it is the eligibility phase-in, income-capped in 2023 and universal by 2025, reaching families who were already in private school. Take out the students who were never going to be in ICCSD and a much smaller number is left for real movement from public to private.

ESA users down to real transfers (2025-26)

Funnel logic from ESA_ICCSD_Decomposition.xlsx. "Upper-bound net move" is the private enrollment above what the pre-ESA private share would predict. "Newly-accredited schools" are Montessori, Hillside, and the Tamarack microschool entering the certified count by getting accredited for ESA. Those students were already non-public.

Why the count overstates the effect

Private share of the resident school-age pool

The private share of the ICCSD-resident pool rose from 7.6% to 9.6% over the ESA period. Real, but modest. Hold the pre-ESA share constant, apply it to each year's resident pool, and the "excess" private enrollment (the part population growth does not explain) is about +311 by 2025-26. That is the upper bound on would-be-public families moving to private. Most of the 1,440 ESA users sit below that line. They were private already.

ICCSD-resident pool (dist 3141)	2022-23	2023-24	2024-25	2025-26
Public certified	14,440	14,379	14,551	14,370
Private enrollment	1,192	1,302	1,434	1,523
Private share of pool	7.6%	8.3%	9.0%	9.6%
ESA residents	n/a	471	773	1,440

Five reasons it holds

The count tracks eligibility, not moves. The 471 to 1,440 ramp follows the phase-in from income-capped to universal, not students changing schools.
Private enrollment grew only a little. Up 331 over three years. The share-adjusted excess is 311, and about 136 of that is schools that simply got accredited.
The public side did not crater. ICCSD public was roughly flat, down 70 across the period. In 2024-25 it actually rose while ESA grew. Public-down does not move with ESA-up.
No new private capacity opened. No new K-12 private school opened in the ESA era. The only physical expansions, Regina's 2020 wing and 2021 early-childhood center, predate it.
The plateau predates ESA. Pre-ESA 2015-16 forecasts overshot 2024-25 by 1,100 to 1,800 students. The Nov-2023 forecast, just after ESA launched, was within about 200. Johnson County births peaked in 2016 and feed smaller kindergarten classes from about 2022 on.

The voucher count is not the cause of the plateau. ESA use is high mainly because families already in private school became eligible for a subsidy. The real move from public to private that ESA can account for is about 120 to 190 students over three years, under 1.5% of the district. It is small, concentrated at Regina, and not what flattened ICCSD enrollment. The plateau is demographic.

What would have changed the conclusion

Two things together would signal a real mix shift. Private schools growing sharply, with new buildings, new K-12 schools, or waitlists. And ICCSD falling below its post-ESA demographic forecast at the same time. Neither happened. The private growth was modest and demographically plausible, and the public count held.

Caveats

District-3141 private enrollment is a close proxy for the ICCSD-resident ESA population, but not identical. Some residents attend private school out of the area, and some boundary-private students are non-residents. Directional, not a census.
The constant-share counterfactual is a simplification, so the page reports a range, not a point.
ESA by school of attendance and open enrollment by receiving district were not available from public sources. Either would tighten the estimate.
A cross-check supports the pool measurement. Public plus private (dist 3141) tracks the ACS school-age population each year, so homeschool and other is small.

What's driving ICCSD enrollment? Full factor decomposition →
ICCSD enrollment forecast, three scenarios →

Source: Iowa DOE Certified Enrollment by District (public & non-public, dist 3141), 2022-23 through 2025-26. Iowa Dept. of Education ESA program counts. ICCSD ESA Private Study (ESA_ICCSD_Decomposition.xlsx, FINDINGS.md). Built iccsd-esa-decomposition.html via scripts/build_esa_decomposition.py.