Fetch NJ Special Education Placement / Educational Environment data
Source:R/sped_placement.R
fetch_sped_placement.RdReturns the IDEA Section 618 "Student Count and Educational Environment"
(placement / Least Restrictive Environment) data published by the NJ DOE,
the companion to fetch_sped (which returns classification
rates). The workbook reports counts and percents of students with
disabilities by educational setting (eg "In General Education for 80
More of the Day", "Separate School", "Residential Facility").
Usage
fetch_sped_placement(
end_year,
age_group = "5-21",
level = "district",
tidy = TRUE,
with_status = FALSE
)Arguments
- end_year
ending school year (eg 2025 for the 2024-25 school year). Valid years: 2020 through 2025.
- age_group
one of
"5-21"(school-age, default) or"3-5"(preschool).- level
one of
"district"(district + charter rows, default) or"state"(statewide breakdowns).level = "school"is rejected: NJ DOE publishes LRE at district/state level only (school-level placement is not_published).- tidy
if
TRUE(default), pivots to the long tidy schema described above. IfFALSE, returns the raw workbook tibble(s) with minimal cleaning (column names preserved as published; all values as character; suppression flags retained). For pre-2025 years that span multiple subgroup files,tidy = FALSEreturns a named list.- with_status
logical, default
FALSE. WhenTRUE(andtidy = TRUE), appends avalue_statuscolumn classifying thecountvalue as"actual"or"suppressed"(a"*"-suppressed small cell). Additive; default output is unchanged.
Coverage
Supports end_years 2020-2025. NJ DOE changed publication conventions multiple times across these years: 2020-2021 are bundled inside an annual zip archive; 2022-2024 publish ~8 single-subgroup workbooks per year; 2025 consolidates everything into one workbook. The fetcher hides these differences and exposes a single tidy schema.
Six state-level slices that NJ DOE published only as PDFs (state 5-21
for end_years 2020-2022, state 3-5 for end_years 2020-2022) are served
from bundled CSVs transcribed from those PDFs; the audit trail (source
URL, SHA-256, transcription date) lives next to the CSVs at
inst/extdata/sped-placement-pdf-transcribed/. The 2023 state 5-21
Excel file ships with NJ DOE's typo "PlacemnetData" in its filename
(StateWide_PlacemnetData_5-21Age_2223_nonpublic.xlsx); the file
map wires this in transparently. Pre-2020 placement data is not
downloadable at all and requires an OPRA request.
Tidy output schema
One row per (entity x subgroup x environment), with:
end_year,county_id,county_name,district_id,district_name(state rows have NA ids anddistrict_name = "New Jersey")subgroup– standardized snake_case ("total","black","hispanic","lep","male", ..., plus disability categories like"autism"and (2025 state output only) age rows like"age_6")subgroup_std– the shared cross-domain subgroup vocabulary (added immediately aftersubgroup). Standard demographic subgroups map; placement-specific / non-demographic tokens ("total", disability categories,age_*,lep,native_american) have no standard demographic equivalent and areNAby design (the row's dimension is given bydimension).environment– short code for the educational setting (see Details for valid values)count,percent– counts and percents (0-100 scale) reported for the cell; suppressed cells ("*") becomeNA. Note: pre-2025 district 5-21 workbooks publish counts only, sopercentisNAin those rows.subgroup_total– the subgroup's row total (Districtwide Total / Statewide Total) carried for convenience. For pre-2025 district 5-21 rows this is the visible-count sum across environments.is_state,is_district,is_charter– entity flags consistent with other njschooldata fetchers (county_id == "80" marks charter schools/districts)dimension(state output only) – which marginal table the row came from ("racial_ethnic","gender","disability","multilingual_learner"; 2025 additionally reports"age")
Environment categories (school-age, 5-21)
gen_ed_80_plus, gen_ed_40_79, gen_ed_less_40,
separate_school, residential_facility,
homebound_hospital, correction_facility.
2025 additionally reports parentally_placed_nonpublic.
Environment categories (preschool, 3-5 state)
ec_program_10plus_hrs,
services_other_loc_attended_ec_10plus_hrs,
ec_program_less_10_hrs,
services_other_loc_attended_ec_less_10_hrs,
separate_class, separate_school, residential_facility,
home, service_provider_location.
The 3-5 district sheet has no environment dimension – the tidy output
uses environment = "districtwide" with the districtwide total.
See also
fetch_sped for the SPED classification rate data,
fetch_sped_placement_multi for a multi-year wrapper, and
get_raw_sped_placement for the underlying raw reader.
Examples
if (FALSE) { # \dontrun{
# 1. Basic call: school-age district-level placement (2025)
placement <- fetch_sped_placement(2025)
# 2. Common dplyr filter -- where are Newark's classified students placed?
library(dplyr)
fetch_sped_placement(2025) %>%
filter(district_name == "Newark Public School District",
subgroup == "total") %>%
select(environment, count, percent) %>%
arrange(desc(percent))
# 3. State-level breakdown by disability (2024)
fetch_sped_placement(2024, level = "state") %>%
filter(dimension == "disability",
environment == "gen_ed_80_plus") %>%
select(subgroup, count, percent) %>%
arrange(desc(percent))
# 4. Earlier-year district 5-21 (pre-2025 returns counts only)
fetch_sped_placement(2022, age_group = "5-21", level = "district")
} # }