Where the numbers on Retraining Report come from
Every program and figure on this site comes from one official file. This page explains what that file is, what we change, and what we leave as reported.
The Department of Labor's list of approved training programs
Under the Workforce Innovation and Opportunity Act (WIOA), every state keeps a list of training programs that workforce funding can pay for. The Department of Labor calls it the state list of eligible training providers; most states call theirs the eligible training provider list, or ETPL. The US Department of Labor publishes all of them together on TrainingProviderResults.gov, with each program's price, length, entry requirements and how its students did. We use the file for program year 2024 (July 2024 to June 2025), which lists 77,085 programs. That makes the data about a year old when it reaches this site, because states report results after students have left.
The list holds only programs whose schools applied for approval. Schools that never applied are missing, so a page never shows everything a state or city offers.
How a program lands on a field, state and city page
Each program carries a federal course code (the Classification of Instructional Programs). Each field of training on this site is one or more of those codes: CDL training, for example, is 49.0205, truck and bus driving, and IT training covers three codes; LPN programs include the code for practical nursing, other, and medical assisting the code for allied health, other. A program filed under one field's code whose name shows it belongs to another is listed there, where people look for it: a Security+ class filed under an IT code is listed under cybersecurity, an LPN-to-RN bridge filed under the LPN code under RN, a phlebotomy or billing course filed under medical assisting under phlebotomy or billing, and a barber or nail course filed under cosmetology under its own field. Certificates, short courses and two-year degrees are listed in every field. Bachelor's degrees are listed only for RN, radiology tech, dental hygiene, sonography, respiratory therapy and medical lab, where a bachelor's and a two-year degree both lead to the same license or certification exam; master's degrees, and degree-completion programs for people who are already licensed (such as RN-to-BSN), are left out. Courses only open to people already working in the job are left out too: cosmetology instructor courses, and fire officer, instructor and fire service management courses under the fire codes (fire academy lists firefighting, fire science and fire protection programs). A program belongs to the state that approved it for funding, even when the school is just across a state line; its card then shows the real city.
City pages cover a whole metro area as the US Census Bureau draws it, so the Dallas page includes Fort Worth and Arlington. A program is placed by the ZIP code of its school. Metro pages list only programs approved by the state the classes are in, and leave out online programs, which appear on the state pages. An area that crosses a state line, such as Washington, DC, holds programs approved by each of those states, and its page says how many by which. An area page is built only when it has at least 10 programs; a state page is built for every field the state has approved programs in, however few.
Some states list the same course several times. We merge listings only when the state, school, city and course name are all identical, so a certificate and a degree at the same school stay separate.
The pages of a state's online programs and of its programs taking 12 weeks or less show up to 10 programs per field, each as a short row, with a link to the field's page, where every program has its full card.
The Department of Labor's results site names the places with "known data quality concerns": Alabama, Colorado, the District of Columbia, Illinois, Louisiana, New Jersey, New York and Puerto Rico (TrainingProviderResults.gov, search page). The pages for those states say so. The same note says Oklahoma has not sent its data for program year 2024, yet the file holds Oklahoma programs; we show them with a notice that their year is uncertain, and will update them when the 2024 data is published. Alabama's column for the credential a program leads to is out of line with its other columns, so Alabama programs are grouped by their names only.
What the price, results and earnings figures mean
- Price is tuition and required fees for someone paying without funding, as the school reported it. Prices under $100 count as not reported, because the file holds many blanks entered as $0. A few prices that are clearly a rate per credit hour, such as $175 for a two-year degree, also show as not reported, and so do a few degree prices far below a single year's tuition, such as a private university bachelor's degree at $1,041 or a placeholder of $99,999. The federal reporting rules ask for the price of the whole program, and that is what the pages show. "Most programs cost" is the middle half of the programs on the page: a quarter cost less and a quarter cost more. Where a page has both short courses and degrees, the price is given separately for certificates and short courses, two-year degrees and bachelor's degrees, because one range over an eight-week course and a two-year degree would describe neither. This happens on every page that is mostly degrees, and on any page where at least two of those levels have five programs with a price.
- Finished is the share of students who completed the program; a share under 1% counts as not reported.Working 6 months later is the share employed in the second quarter after they left. Where a state's figures look wrong as a whole, we leave that figure out for the state rather than print it: the share working 6 months later for Alabama, Idaho, Louisiana, Nevada, New Hampshire, Oklahoma, Pennsylvania, Puerto Rico, South Dakota, Utah, West Virginia and Wyoming (rates far below every other state's, or far too many at exactly 100%), and the share who finished for Illinois, Louisiana, Mississippi, New Mexico and Puerto Rico (in all but Illinois, half or more of the programs report exactly 100%). Each of those states' pages says which figure is missing and why. Where one school's figure cannot be right, such as a share working of about 5% next to high earnings, we leave out that figure alone, and the state's page names the school and the reason.
- Earnings are the middle earnings of those working in that second quarter. States take them from the wage reports employers file every three months, so the file has one quarter. The yearly figure is that quarter times four, rounded to the nearest $100, and marked as an estimate. A quarter under $1,000 counts as not reported.
- Trained with WIOA funding counts the people in a program whose training WIOA paid toward through a training account in the program year. The file never gives counts of 1 or 2, so a program without the figure may still take WIOA students, and a page's total is "at least". It does not say how much of each price WIOA covered: the dollar amounts in the file come to a small fraction of the price in whole states and in some local areas, so they are not shown.
- Length is the program's length in weeks as reported. For CDL, CNA and phlebotomy courses, a length over two years counts as not reported; for medical assisting, dental assisting, pharmacy technician, auto mechanic, cybersecurity, IT training, project management and accounting programs, a length over three years; for single classes, short courses and pre-apprenticeships in any trade, a length over two years. Apprenticeships keep their four or five years. A two-year or bachelor's degree filed as lasting under 30 weeks counts as not reported: no degree takes that little time, so the figure is most likely one term. A length of exactly 260 weeks, which some states enter when a program has no set length, also counts as not reported.
A typical figure for a whole page, such as how many students finished, is the middle program among those that report it: half did better, half did worse. Such a figure, like what most programs cost, is shown only when at least 5 programs report it and they make up at least a quarter of the page. A page with one or two programs shows those programs' own figures in its place. A larger page with too few reports says "Not enough data"; each program's own figures are on its card. A page says "Not reported" only when none of its programs report the figure, and "Left out" when the only figures filed were removed by one of the checks above.
Not reported, left out or not enough data: why a figure is missing
Each of the three has its own small mark, so you can tell them apart at a glance.
- Not reported: the state or school left the field empty, entered $0 as a price, or the Department of Labor withheld the figure to protect small groups of students.
- Left out: the file holds a figure, but it fails one of the checks on this page: a price under $100, a length no program of its kind takes, earnings under $1,000 a quarter, a completion rate under 1%, a rate a whole state reports in a way that can't be right, or one school's figure that can't be right. Anyone who opens the source file will find a value there; the state's page names whole-state and single-school cases.
- Not enough data: some programs report the figure, but too few for a typical figure for the whole page; each program's own figure is on its card.
In the comparison tables, a row with no figures at all says so once: "No figures reported" when none of its programs report any, "Too few figures to show" when some do, but too few for a typical figure.
Which schools we mark as accredited, and how we checked
A school is marked "Accredited school" when it is on the US Department of Education's Database of Accredited Postsecondary Institutions and Programs with current accreditation (accredited or preaccredited, with no end date). We use the Department's data files of October 24, 2026. The Department doesn't accredit schools itself: it reviews accrediting agencies and lists the ones it recognizes, along with the state agencies it recognizes for public vocational and nurse education. It also notes that the agencies report this information, that it isn't audited, and that the agency has the latest status.
The Department's list and the Department of Labor's file share no ID number, so we matched them. A school matches when its website is the one the Department's college directory (IPEDS, 2024-25) gives for an accredited institution in the same state, or when its name is the same as an accredited institution's or campus's in that state. Every other likely pair, about 2,300 schools, was checked by hand, and a school we couldn't place with confidence stays unmarked. A campus counts under the institution it belongs to. Colleges also list the places they teach at, such as high school career centers, hospitals and prisons; a school that only hosts a college's classes doesn't count as the college. About 2,900 of the 7,500 schools on the site match, and they hold about 72% of the programs.
Each accredited school's row names its accreditor, the year it was first accredited where the Department marks that date as actual, and its next review. Where the list also records accreditation of a program at that school, such as an associate degree nursing program, the row names it under the matching field. The Commission on Accreditation of Allied Health Education Programs (CAAHEP) accredits programs in medical assisting, sonography, surgical technology and other fields, but its directory may not be reused without its permission, so those pages link to CAAHEP's own search instead.
A state gets a page of accredited schools for a line of work once 5 accredited schools there offer programs in it. Accreditation isn't a condition of WIOA funding, and many short courses come from schools that aren't on the Department's list; the mark says only that a school is on it.
What Retraining Report never changes or adds
Figures for named schools are shown exactly as reported, without labels like "best" or "worst". Where a state left a field empty or the Department of Labor withheld it to protect small groups of students, the page says "Not reported"; nothing is estimated to fill the gap. When a school listed the same course twice with different results, the card shows the listing with the most students and says so. The file also records whether a school is public or private, but those entries are wrong too often to print, so we leave them out. A few web addresses in the file cannot open because the slash after the school's domain is missing; those cards link to the school's home page and say "School website".
Facts that don't come from the file, such as who qualifies for funding, financial aid and the GI Bill, link to official sources, numbered at the bottom of each page.
How to use these figures when choosing a training program
Results describe students who left a program in recent years. They are no guarantee of how the next class will do, and prices change. Before you enroll, confirm the price, length and funding with the school and your local American Job Center. If something here looks wrong, tell us and we'll check it against the source.
Last updated 25 October 2026