How many Lunenburg children are on an IEP

What this establishes

The count, year by year

The paraprofessional figure, and why only one row of it is published

What the children are, by DESE’s own categories

Where they are taught — and what that table leaves out

A third file, and a fourth answer

What the town said about this

The documents

What this report cannot answer

The other three reports

Analyses

258 of them, in FY26. The state publishes the count; this is what it says.

lunenburgbudgetproject.org — written by the Lunenburg Budget Project, an independent tool for residents. Not affiliated with the Town of Lunenburg, the School Committee or the school district.

258

children on an individual education programme in FY26

16.3%

of the 1,581 students DESE counts as enrolled

246

educated in Lunenburg

placed out of district

What this report counts

Children, counted by the state at its own census. Nothing on this page is a dollar, an FTE budget line, or a cost per pupil — those are a separate report, and they are separate because a student is not a dollar and every retraction this project has had to make came from a sentence that slid between the two.

Finding 1

The count is up 31 since FY22, and the share is up 2.2 points

227 children in FY22, 258 in FY26, against a total enrolment that went from 1,608 to 1,581. So the proportion moved from 14.1% to 16.3%. This is DESE’s own published count, not a percentage multiplied back out by us.

Finding 2

The two published parts add up, in every year

DESE prints a total and prints in-district and out-of-district separately. They reconcile exactly in all 6 years. That matters because the staffing rates in the same file are computed on the in-district figure, which is 12 lower than the count printed beside them — a disagreement this project had registered as unexplained. It is not unexplained. The difference is the out-of-district count, exactly, in every year.

Finding 3

The state's paraprofessional FTE for Lunenburg falls 63% while the count of children does not

49 FTE in FY19, 18 in FY26. Per 100 students with disabilities that is 18.5 falling to 7.3. The denominator over the same span went from 265 to 246, so this is not a base effect.

Finding 4

Only 1 of the 4 staff rows in that table reproduces from its own printed figures

Every row prints an FTE, a count and a rate. Recomputing the rate from the other two: paraprofessionals 8 of 8; service FTE 0 of 7; support FTE 0 of 6; teachers 1 of 8. So this page publishes the paraprofessional row and refuses to publish service FTE, support FTE, teachers as figures at all.

In district and out of district, from DESE’s Special Education Program Characteristics file. The two segments are the file’s own split and they sum to the total it prints.

educated in Lunenburgplaced out of district

6 years · FY2021–FY2026 · DESE’s published count of students with disabilities

the same rows, as numbers

year	on an IEP	in district	out of district	enrolled	share
FY21	261	249	12	1,605	16.3%
FY22	227	217	10	1,608	14.1%
FY23	234	230	4	1,567	14.9%
FY24	233	224	9	1,589	14.7%
FY25	248	238	10	1,573	15.8%
FY26	258	246	12	1,581	16.3%

What this does not show

That a rising count means rising need. A count of children with an IEP is a count of children who have one. Referral rates, evaluation practice, eligibility decisions and the arrival of a family who already had a plan all move it, and none of those is published. A count is a real quantity and it is not a measure of how many children require services.

What this does not show

What happened before FY21. This file starts there. A longer series of the out-of-district half alone exists in the town’s own annual reports and is on the route report — on its own frame, because it is a different instrument measured on a different date.

DESE reports special education staffing as a rate per 100 students with disabilities, and prints the FTE and the denominator beside the rate. That makes every row checkable: the rate has to reproduce. Recomputed on every build, the paraprofessional row reproduces in 8 of 8 years and the others do not, which is why the chart below has one line for staff and not four.

special education paraprofessional FTE (left axis)students with disabilities, the return’s own denominator (right axis)

8 years · FY2019–FY2026 · DESE Special Education Indicators — the one staff row that reconciles

the row, and its own arithmetic

year	FTE printed	students printed	per 100	recomputed
FY19	49	265	18.5	18.5
FY20	44	244	18.0	18.0
FY21	41.5	249	16.7	16.7
FY22	32.5	217	15.0	15.0
FY23	24	230	10.4	10.4
FY24	23	224	10.3	10.3
FY25	20	238	8.4	8.4
FY26	18	246	7.3	7.3

every staff row in the same table, and whether it reproduces

row	reproduces in	years published
Special education paraprofessionals per 100 SWD	8	8
Special education service FTE per 100 SWD	0	7
Special education support FTE per 100 SWD	0	6
Special education teachers per 100 SWD	1	8

What this does not show

That this is a count of posts. The registry already records two reasons it cannot be: DESE’s FTE times the published count implies a per-FTE figure a paraprofessional does not cost, so the state’s figure and the district’s budget line are not describing the same population; and DESE’s special education TEACHER FTE for Lunenburg falls about 90% while total teacher FTE holds flat, which is the signature recoding would produce. A fall in this series is consistent with fewer posts, with the same posts reported differently, and with a change in how a part-time assignment is apportioned.

A possible explanation — nothing here tests it

The federal grant money that ended in FY2025 was paying for staff in the function DESE calls Paraprofessionals — measured on its own page, where grants in that function fall and the general fund rises in the same year. Two series moving the same way is not one series explaining the other, and the years do not line up cleanly: this fall begins before the grant fall and continues after it.

What would settle it: EPIMS work assignment records for Lunenburg by job classification and year, which would show whether the same people were reported under a different code.

“they are the backbone of special education and without them the department cannot run”

School Committee · 2024-01-24 · our copy · the town’s

A parent, in the public comment on the FY25 budget, on the proposed paraprofessional cuts. Said in the year between the two largest falls in the state’s own paraprofessional FTE figure for Lunenburg. It establishes that people knew posts were at stake. It does not establish what the FTE return counts, which is the open question beside it.

“Ms. Brzozoski makes a motion to add two paraprofessional positions where needed, Ms Young seconds the motion, all approve”

School Committee · 2025-10-01 · our copy · the town’s

A vote to ADD posts, in the same school year DESE’s return reports the lowest paraprofessional FTE in the eight years it covers. Both are real. A vote to add positions and a state FTE return are different quantities measured by different people at different moments, and nothing here reconciles them.

The uncollapsed disability categories. Between FY21 and FY26 the largest movement up is Specific Learning Disabilities (+8.7 points of the total) and the largest down is Developmental Delay (-6.3 points). Those are movements in a classification, and a classification is a decision somebody made about a child rather than a fact about the child.

share of the count, by disability category, FY2021–FY2026

category	FY21	FY22	FY23	FY24	FY25	FY26
Autism	47 · 18.0%	42 · 18.5%	42 · 17.9%	56 · 24.0%	53 · 21.4%	48 · 18.6%
Communication Disability	28 · 10.7%	24 · 10.6%	25 · 10.7%	19 · 8.2%	17 · 6.8%	18 · 7.0%
Developmental Delay	59 · 22.6%	49 · 21.6%	42 · 17.9%	24 · 10.3%	36 · 14.5%	42 · 16.3%
Emotional Disability	23 · 8.8%	19 · 8.4%	21 · 9.0%	17 · 7.3%	18 · 7.3%	18 · 7.0%
Health Impairment	37 · 14.2%	31 · 13.7%	30 · 12.8%	37 · 15.9%	40 · 16.1%	44 · 17.1%
Intellectual Disability	7 · 2.7%	5 · 2.2%	4 · 1.7%	5 · 2.1%	5 · 2.0%	8 · 3.1%
Multiple Disabilities	15 · 5.8%	12 · 5.3%	11 · 4.7%	13 · 5.6%	13 · 5.2%	12 · 4.7%
Neurological Disability	8 · 3.1%	9 · 4.0%	11 · 4.7%	9 · 3.9%	10 · 4.0%	9 · 3.5%
Sensory Impairment	1 · 0.4%	1 · 0.4%	2 · 0.8%	1 · 0.4%	—	1 · 0.4%
Specific Learning Disabilities	36 · 13.8%	35 · 15.4%	46 · 19.7%	52 · 22.3%	56 · 22.6%	58 · 22.5%

What this does not show

That the mix shifting means anything about severity, cost or service intensity. CLAUDE.md carries “the mix shifted toward intensity” as one of the sentences this project shipped as fact when nothing established it. A category is an eligibility label. Nothing here links a label to an hour of service or to a dollar, and the categories themselves were not defined by Lunenburg.

DESE publishes a Placement breakdown in the same file. It is in-district only, and its parts run short of the total printed beside them. In FY26 the parts sum to 242 against a published total of 258 and an in-district count of 246. The shortfall is shown rather than the rows being rescaled to look complete.

placement, FY2026

setting	children	share of the published total
Full Inclusion	183	70.9%
Partial Inclusion	16	6.2%
Separate School in District	2	0.8%
Substantially Separate	41	15.9%

the parts against the totals, every year

year	parts sum to	in district	published total	unaccounted
FY21	246	249	261	15
FY22	214	217	227	13
FY23	229	230	234	5
FY24	224	224	233	9
FY25	235	238	248	13
FY26	242	246	258	16

by grade span

What this does not show

Which children the unaccounted rows are. The category is documented as in-district only, which explains part of the shortfall and not all of it — in FY26 the parts fall short of even the in-district count. Suppression of small cells fits, and so does a setting the table does not name.

DESE also publishes a caseload movement file. Its count of children on an IEP is lower than the count above in every overlapping year — by 26 in FY25, the widest — and it reports a different enrolment to go with it. Both are DESE, both are Lunenburg, both are the same school year.

two DESE counts of the same thing

year	movement file	programme file	difference
FY21	241	261	-20
FY22	207	227	-20
FY23	224	234	-10
FY24	222	233	-11
FY25	222	248	-26

what the movement file is for — children entering and leaving services

year	enrolled	on an IEP	moved in	moved out	note
FY19	1,518	254	28	40
FY20	1,477	242	24	19
FY21	1,437	241	19	28
FY22	1,439	207	33	28
FY23	1,438	224	23	21
FY24	1,448	222	38	17
FY25	1,448	222	38	17	identical to the prior year — do not read as a second year

What this does not show

Which of the two is right, or what they count differently. An October census against a March one, a K-12 grade basis against an all-grades one, and the inclusion or exclusion of out-of-district children all fit. Note also that FY25 repeats the previous year across every column in this file, which the extract flags in the data itself so that two years of movement are never read where one was published.

Rule 15a. The measurement above is the state’s; this is what was said out loud in the same years, and none of it is a measurement.

documents mention “special education”

documents mention “IEP”

documents mention “paraprofessional”

documents mention “students with disabilities”

Searched 8,899 of the 12,015 meeting documents this archive holds (74%), covering 2009-01-05 to 2026-11-17. The other 3,116 carry no text a search can match — 3,096 of them are image scans awaiting OCR. An empty result above is a statement about what can be read, never about what was said.

dese-sped-program-characteristics.xlsx

Massachusetts Department of Elementary and Secondary Education

Special Education Program Characteristics and Student Demographics (n62c-bx65). Carries the published count itself, not a percentage.

table dese_sped_program · 5.4 MB

sha256 417435bddfcac3eb…

our copy · the publisher’s

dese-sped-indicators.xlsx

Massachusetts Department of Elementary and Secondary Education

Special Education Indicators (yamx-769q). The staffing rates per 100 students with disabilities.

table dese_sped_indicator · 12.1 MB

sha256 41d517465d5923be…

our copy · the publisher’s

dese-sped-movement.xlsx

Massachusetts Department of Elementary and Secondary Education

Students Moving In and Out of Special Education Services (8aww-sugs). A THIRD count, on a different basis.

table dese_sped_movement · 1.1 MB

sha256 2b5e67f3ba6b1cbd…

our copy · the publisher’s

That a rising count means rising need. A count of children with an IEP is a count of children who have one. Referral, evaluation and eligibility practice all move it, and none of them is published here.
What the special education FTE return counts. Only the paraprofessional row reproduces from its own printed FTE and count; the teacher, service and support rows in the same table do not, in any year, so they are not published as figures on this page.
That the paraprofessional FTE series is a count of posts. `money-gaps.csv` already records that DESE’s FTE and the district’s paraprofessional budget line cannot describe the same population, and that DESE’s special education teacher FTE for Lunenburg falls 90% while total teacher FTE holds flat — a signature that recoding would produce exactly.
How the third count relates to the other two. `dese_sped_movement` reports a K-12 grade-row basis and a different enrolment; nothing published states the census date or the inclusion rule behind either.
A breakdown of every child by setting. DESE’s Placement category is in-district only and its parts run short of the total printed beside them; the shortfall is published on this page rather than hidden by rescaling.

What would close these

DESE’s SIMS submission detail for Lunenburg, stating the census date and the inclusion rule behind each published count, and the EPIMS work assignment records that would say what a special education FTE is a count of. Every limit on this page is also a row in what we cannot answer, which is the single registry the records request reads from.

Grades 3 to 5

Grades 6 to 8

Grades 9 to 12

Special Education Beyond Grade 12

Four reports, deliberately not one. Each counts something different, and the difference is the point — a placement is not a cost and a dollar is not a child.

Who leaves, and where they go →counts children, with no disability flag What it costs, and what comes back →counts dollars The route out of district →counts placements