Every change to a published number, method or data source is listed here, newest first: what changed, why, how large the effect is, and which pages it touched. Corrections are listed even when the effect is small. Pages changed since November 9, 2026 also note the change in their own update line. To report an error, write to info@opportunitydata.org.
Correction
Workforce Pell for Students: the Student Aid Index ceiling
The students page and the advising-office guide said Pell has no fixed income cutoff. Since July 1, 2026, a student whose Student Aid Index equals or exceeds twice the maximum Pell Grant for the year is not eligible (Higher Education Act section 401(b)(1)(F), added by Public Law 119-21, section 83001(b)). Both now state the rule.
The same review corrected how the students page describes approval: 22 of its 30 programs have a U.S. Department of Education announcement and 8 are listed as federally approved on official state rosters, and two Eastern Florida State College programs begin taking Workforce Pell in 2027. Wording that every program could be used “right now” was removed.
The Class of 2026 job market: AI attribution and the low-unemployment summary
On November 10, a sentence that attributed elevated recent-graduate unemployment to AI on the New York Fed’s authority was corrected; the New York Fed series assigns no cause. On November 12, a second unsourced attribution to AI, in the computer-fields section, was removed.
The summary said nursing was the one low-unemployment field paying above the recent-graduate median. The page’s own table shows aerospace (2.2 percent unemployment, $85,000) and civil engineering (2.3 percent, $75,000) above the median as well; the summary now names them.
The page now states that its field-level CSV and methodology note use 2022-23 completion counts, while the page uses 2023-24. No rate changed.
State Implementation Record: New Jersey program count
The Record’s summary said New Jersey listed three programs at the Governor-recommendation stage. The state page lists four, and the Record’s New Jersey row already said four; the summary now names the fourth, AIMS Education’s Medical Billing and Coding Specialist.
Opportunity Data AI Exposure Index: occupation titles for eight “All Other” codes
The September 2026 data release of the Opportunity Data AI Exposure Index labeled eight SOC 2018 codes with the title of an O*NET specialty instead of the BLS occupation title. SOC 19-3099, for example, appeared as “Transportation Planners”; the BLS title is “Social Scientists and Related Workers, All Other.” The other seven are 15-2099, 19-2099, 25-2059, 29-9099, 33-9099, 49-9099 and 51-8099. A reader reported the error.
For each of the eight codes, O*NET rates only one specialty within the group (for 19-3099, Transportation Planners, 19-3099.01), so the score for the whole code is computed from that specialty’s profile. That scoring rule is unchanged from the April 2026 release, which used the BLS titles. The occupation explorer now shows the BLS title and names the O*NET specialty each of the eight scores comes from.
No score, employment figure or program result changed.
Workforce Pell value-added earnings: price parity follows each college’s location
The value-added earnings test divides median earnings by a regional price parity before subtracting 150 percent of the poverty line. The rule, 34 CFR 690.95(b)(1), assigns the parity “based on the location of such programs”: the parity of the metropolitan area a program is in, or the state parity when the program is outside a metropolitan area. The final rule’s preamble (91 FR 29254, May 19, 2026) rejects using BEA’s nonmetropolitan-portion value instead.
The value-added calculator used the state parity for every college. The Colorado and North Carolina program lookups used metropolitan parities but gave colleges outside metropolitan areas the nonmetropolitan-portion value (97.5 in Colorado, 88.4 in North Carolina; the state values are 103.1 and 94.3), and the North Carolina lookup did not use the published Pinehurst-Southern Pines parity (88.3). All three now follow the rule, using BEA’s 2024 parities (released February 19, 2026) and each college’s location in IPEDS 2024.
Effect in the calculator: of 3,619 certificate programs with published first-year earnings and a tuition figure, 132 change their result against annual tuition, 83 to clearing it and 49 to falling short, and 28 change whether value-added earnings are positive at all. Colleges in high-price metropolitan areas now show lower value-added earnings, and colleges in lower-price areas higher ones. Of the calculator’s 602 colleges, 48 have no location match and keep the state parity, which the page states for each one. In the lookups, 124 Colorado and 554 North Carolina program rows show a new parity.
Not affected: the Earnings Accountability Record. The degree-program earnings test in 34 CFR 668.403 has no price parity adjustment.
The site now says that Census PSEO 2026Q2 publishes earnings for 1,101 institutions at the three degree levels the site covers, in 37 states and the District of Columbia. It had said 38 states, counting the District as a state. The earnings hub still carried the April figures (938 institutions, 33 states). Bachelor’s coverage is 36 states and the District, not 37. The home and About pages now say more than 1,100 institutions, not more than 900, and the coverage map’s state coloring, missing since July 15, is restored.
State price parities replaced with BEA’s 2024 values
The state price-parity table behind the value-added calculator and the community college baccalaureate comparison was labeled 2023 but matched no BEA release: 0 of 51 states agreed with the 2023 values as revised and 9 of 51 with 2024. Tennessee, for example, was 90.5, Kentucky’s 2023 value; Tennessee’s 2024 value is 91.9. The table now holds BEA’s 2024 state values (all items, released February 19, 2026).
Value-added calculator: first-year earnings by default
34 CFR 690.91 measures earnings in the first full tax year after the award year of completion, so the calculator now shows Census PSEO year-1 earnings by default, the closest public match. Year-5 earnings stay available for comparison. The page previously described the proposed rule; it now cites the final rule.
This Week: the Department’s proposed supplemental priorities
The November 4 item said the Department’s proposed supplemental priorities (91 FR 69736, comments due December 7) add Workforce Pell activities. They do not. The final Career Pathways priority of April 13 (91 FR 18780) added them, and the proposal renumbers them. The item and the key-dates entry are corrected, and an April 13 entry is added.
California Workforce Pell list: earnings test with the price parity applied
The California analysis compared wages with the value-added floor without dividing by California’s price parity (110.7), which 34 CFR 690.95(b)(1) requires. Corrected, all 546 listed occupations still clear the test at $3,000 in tuition, and 544 of 546 clear it at $9,000 (it had said all 546). Child development misses $9,000 on completer earnings. A metropolitan rerun, pairing each area’s wages with its own parity, finds no failures at $3,000 anywhere.
Child care and teacher credential ladders: Census degree codes
The child care, Colorado child care, teacher and healthcare pathway pages read Census PSEO degree codes 02, 03 and 05 as associate’s, bachelor’s and master’s. In the Census label file, 02 is a one-to-two-year certificate, 03 an associate’s degree and 05 a bachelor’s degree. With the codes corrected, child development certificates and associate’s degrees both start near $27,600, and bachelor’s graduates start at $33,556 and reach $47,122 by year five; the pages had shown no bachelor’s premium. Master’s degrees have no published PSEO earnings at this level, so the teacher page no longer compares them.