A Credential Is Not Evidence That the Move Pays
Completion is not employment, and employment is not a durable wage gain. Put a retraining programme through a four-part evidence test.
Reskilling supplies the programme outputs and outcome measures; career-transitions supplies the occupational destination, employer demand, and wage-and-access costs that make those measures matter.
A strong evaluation does not create an opening for an older worker, a person without time for unpaid practice, or an outsider entering a crowded destination without its network.
The payment page is open on your phone.
The countdown clock is moving.
The course has a logo, a public funding badge, a promise from an employer, and a photograph of people holding certificates.
Your layoff email is still open behind it.
Maybe the offer is a bootcamp. Maybe it is a public programme. Maybe an old employer says the degree will make a redeployment possible.
Do not decide whether the programme is respectable.
Decide whether the move pays.
The certificate is the easiest thing for a provider to count.
The job is harder.
The wage is harder still.
The cost of getting from this week to that wage is where the sales page goes quiet.
Do not send the deposit while the silence is doing the work.
The Output
A credential says you attended, completed, or met an assessed learning standard.
It does not say you entered the intended occupation.
It does not say the job lasted.
It does not say the wage repaired your loss.
Keep the nouns separate.
Completion is a programme output.
Employment is an outcome only after somebody says what counts as employment and when it is measured.
Wage recovery is a further outcome.
A durable wage gain needs a comparison with what would otherwise have happened to you.
The TAA report gives the category its cleanest warning.
Nearly 90% of people who completed training received an industry-recognised or secondary credential.
That is a high figure for a thing that can be issued at the end of a programme.
It is not a causal wage return.
Do not let a provider move from one noun to the next without evidence.
Certificate.
Placement.
Employment.
Retention.
Earnings.
Those are not five names for the same event.
They are five places where your transition can fail.
Four Answers
Take out a sheet of paper.
Write the provider’s name at the top.
Write four questions underneath.
Do not ask whether graduates get hired.
That question has too many exits.
Ask these.
What is the completion denominator? Ask how many people were recruited, enrolled, started, completed, and received the credential. Ask for cohort dates and exclusions. A programme can advertise completion among people who survived intake, began the course, or stayed until the final assessment. Each denominator tells a different story.
What counts as employment, and when? Ask for the intended occupation, start date, employer, hours, contract type, wage, benefits, and follow-up date. A paid placement, a temporary contract, or an unrelated income job can matter to your household. None of them proves an occupational transition.
What happened to wages? Ask for the median and distribution of post-programme wages against pre-programme wages, not the highest graduate salary. Ask how many people regained their prior hours. Ask what happened after the first paycheque. Then ask what comparison exists for people who did not take the programme.
Which employers are the destinations? Ask for names, not a sector collage. Ask which entry jobs they hire for, what vacancies exist now, what screening they use, and whether the provider has more than one employer in the route.
Score one point for each answer that is independently documented.
An answer on a sales page is a claim.
An answer in a cohort report with dates, a method, and a denominator is evidence you can inspect.
4 points: proceed. The route has supplied the minimum evidence you need to examine cost, fit, and current demand. Paying is still your decision. The provider has merely earned the right to be considered.
2–3 points: investigate. The missing answer may be recoverable. Email the questions. Check the occupation, the employers, and your cash calendar before you sign anything.
0–1 point: do not pay. The provider is asking you to carry the unknown parts of its business model.
Four questions cannot be validated against anything.
They are not a rating system.
Their purpose is smaller and more useful.
They stop a certificate from impersonating a transition.
The Denominator
Start with completion.
That is where a clean percentage usually conceals a messy route.
Ask who had to leave before instruction began.
Ask who began and did not finish.
Ask who took a different job because rent arrived before graduation.
Ask who lost transport, child care, equipment, internet access, or the hours required for unpaid practice.
A serious intake plan names the target occupation, prerequisite work, start date, attendance plan, support referrals, and a date for reconsidering the target.
That list exists because ready is not a fact.
It is a bundle of conditions.
Pre-training workshops in the HPOG study were associated with a greater likelihood of both starting and completing health-care training.
Association is not proof that a workshop causes completion.
It does show something you need.
The people who reach a programme’s certificate count are not interchangeable with everyone who first needed a new occupation.
The programme must tell you where the count begins.
Do not accept 90% completion until it tells you 90% of whom.
The Destination Screen
Employment needs a destination screen.
Write down your origin occupation.
The target occupation.
The work region, the required licence or credential, and the entry title.
The wage and schedule.
The employers, and the realistic next role.
That is not paperwork.
That is the route.
Career-transition research separates a direct job-to-job move, a move after unemployment, internal redeployment, and a labour-force exit.
A temporary spell outside work cannot be quietly folded into a direct transition.
BLS’s permanent-separation rule treats a labour-force exit as at least four consecutive months out, and an occupational transfer as a move to a different major occupational group.
That is one stated definition, not a census of all career movement.
Use the discipline, not the rule, when you read a provider’s placement number.
What is the occupational classification?
What is the observation window?
What happens to graduates with several short jobs, an unknown destination, or no work at the follow-up date?
They are not noise.
They are part of the denominator the provider wants you to forget.
Your occupational move also needs current demand.
A frequent historical route can lead into another shrinking labour market.
An employer advisory board can contain no actual hiring promise.
An employer talk can end before a resume is opened.
Reject a destination with no local hiring evidence, even when the course brand is famous.
The route must name the employer, vacancy, eligibility, wage, supervisor, assessment or paid work trial, offer, start date, and the person responsible when you hit an early barrier.
A referral to training or a job board is not a transition pathway.
It is a handoff without a destination.
The First Placement
The provider may say graduates are employed within three months.
The TAA report recorded 74% employed within three months after receiving benefits and services.
Keep the qualifiers attached.
It is a descriptive outcome from that programme.
It is not a target that makes unrelated work, unstable hours, or lower pay a successful retraining result.
The provider must say whether the work is in the target occupation.
It must say whether the person is still there.
Your first 30 to 180 days after placement are where a move comes undone.
Transport.
Schedule.
Equipment, onboarding paperwork, workplace conflict, and an unexpected skill gap.
Measure the first job.
Then measure employment, occupation, employer, hours, pay, and retention again.
Year Up makes the point without offering comfort.
At four years, employment was 82% for programme members and 80% for controls.
Employment alone did not settle the question.
Job quality and earnings did the work a binary employment figure could not.
A destination with six-month retention and persistent wage loss is a bridge route.
Call it that.
Do not call it recovery.
The Before Picture
Write down your wage and hours before the layoff.
Write down the replacement wage and hours the course says are likely.
Then write the months with no or reduced earnings between them.
Do not treat a graduate’s wage level as a return on investment.
If you would have found comparable work without the programme, the provider did not create the whole result.
If you work more hours at a lower rate, a headline wage can hide the loss.
If you move into an occupation with delayed progression, the first year can hide the later gain.
Compare outcomes with a matched group not exposed to the programme, where that evidence exists.
Per Scholas gives a more serious example than a testimonial.
In the WorkAdvance evaluation, follow-up earnings exceeded controls by $5,186, with the difference emerging after the fourth quarter.
That is useful because it compares a programme group with controls and says when the difference appeared.
It still does not prove a break-even date.
The estimate does not settle taxes, benefits, participant costs, different time horizons, or whether another provider can reproduce the result.
Use one, three, or five years as a stated horizon.
Subtract tuition.
Subtract examination fees, tools, travel, care, data access, debt cost, taxes, lost benefits, and foregone earnings.
Then ask what would have happened without the course.
That last question is not pessimism.
It is the counterfactual.
Without it, the provider compares graduation with nothing and calls the difference success.
The Real Cost
The fee is displayed in large type because it is the amount the provider wants paid now.
Your cost is larger, because your time is not free.
In OECD experimental estimates, opportunity cost averages about 75% of total per-person transition cost across occupations and scenarios.
The calculation multiplies origin-occupation median wage by training duration.
That is not a bill that arrives in the post.
It is a warning to count income you do not receive.
In the Greece evaluation, the lock-in period was roughly four months, approximately the programme duration.
That is the period your wage comparison must carry, not erase.
OECD modelled average total transition costs at about USD 7,000 for routes requiring up to six months of training, USD 13,000 for six months to a year, and USD 50,000 for one to three years.
These are modelled averages across countries.
They are not a tuition menu.
They do tell you why a public programme, an employer promise, or a degree can be economically different from its price tag.
Build two calendars.
Put fees, exam costs, equipment, travel, care, and relocation deposits on the cash calendar.
Put lost wages and reduced hours on the income calendar.
Put benefits, stipends, wage insurance, employer pay, and tax credits on the date they actually arrive.
Two routes with the same total cost can fail differently when your rent is due before the wage lift.
Free tuition is not free access.
The Provider’s Ledger
Do not use a provider’s operating cost as proof that its graduates come out ahead.
In the four-site WorkAdvance evaluation, in-program operating expenditure ranged from $5,200 to $6,700 per participant.
Those figures include more than instruction.
Recruitment, screening, coaches, facilities, and equipment.
Assessment, employer engagement, placement, administration, and support.
That is why a low-tuition offer may be withholding the parts that make a route possible.
It is not proof that a higher-fee offer is better.
The same evaluation estimated net costs of about $3,500 per participant at Per Scholas, and $4,900 to $5,900 at the other three sites, after accounting for alternative services.
Gross cost asks what it takes to run the provider.
Net cost asks what extra spending exists because this alternative was offered.
Neither asks whether you recovered your prior earnings.
Keep the payer visible.
You.
The provider.
The employer.
The public funder.
Your household.
The party that advertises the return is rarely the party carrying the risk.
The Employer Promise
An employer-sponsored conversion can reduce demand uncertainty.
It does not remove it.
Ask for the wage, hiring criteria, number of openings, entry title, geography, supervisor, and what happens if demand falls.
Ask whether an interview is reserved, a placement is reserved, or the employer has merely agreed to consider graduates.
Those are different agreements.
WorkAdvance combined four linked components rather than treating classroom instruction and placement as separate products.
Year Up paired six months of training with a six-month paid internship.
Those designs show why a course can be more than a lecture series.
They do not turn a provider’s employer logo wall into evidence.
Check concentration too.
One employer can pause hiring.
One contract can end.
One local labour market can fill with people carrying the same new credential you just bought.
Your destination needs a portfolio of employers.
Not a single photograph from a graduation day.
Missing Evidence
The certificate is not fraudulent because it is incomplete.
It is a provider’s record of attendance or assessed learning.
The problem begins when that record is sold to you as proof of a move nobody has measured.
Missing employment evidence tells you how much risk the provider expects you to carry.
Missing wage evidence tells you whether the provider is measuring a recovery or merely a start.
Missing employer names tell you whether the destination exists outside a brochure.
Strong evidence still does not guarantee you access.
If you are older, you may meet a crowded entry market after the course ends.
If you have no time for unpaid practice, a pathway advertised as open can exclude you.
If you come from outside the destination’s network, you may face screening rules the course cannot change.
Do not turn that into a personal defect.
Put it in the cost.
Put it in the destination screen.
Then return to the four points.
A countdown clock is a pricing device.
It is not evidence about a labour market.
It will still be there tomorrow under a different number.
Minimum Viable Action
Before you sleep tonight, email one provider the four questions.
Ask for the completion denominator.
Ask what employment means and when it is measured.
Ask for post-programme wages against pre-programme wages, and the comparison used.
Ask for named destination employers.
Do not send money until it replies.
What you will be able to say afterwards
- A completion rate needs a stated denominator before it can be used as evidence.
- Employment needs a stated occupation, timing, hours, and retention rule before it counts as an outcome.
- Post-programme wages need comparison with pre-programme wages and, where possible, with a credible counterfactual.
- Named destination employers and current vacancy evidence test the occupation rather than the course brand.
- Tuition, foregone earnings, transport, care, equipment, and delayed income belong in the transition cost.
- A four-point retraining test separates proceed, investigate, and do-not-pay decisions.
Questions this raises
Does a certificate prove that a retraining programme works?
No. A certificate records attendance, completion, or assessed learning under a provider's rule. It does not establish that graduates enter the intended occupation, keep work, recover prior earnings, or obtain a durable wage gain. The TAA report's nearly 90% credential figure is an output measure. It is not causal evidence that every credential produces a wage return.
What has to be in a completion rate before it means anything?
A completion rate needs its denominator. A provider should state whether it counts everyone recruited, everyone enrolled, everyone who started instruction, or only people still present near the end. It should also state the cohort dates, the credential rule, and who is excluded. Without that denominator, a high completion percentage can describe selection or disappearance rather than programme performance.
What counts as employment after a course?
Employment needs a stated definition: the intended occupation, employer, start date, hours, wage, contract type, benefits, and the follow-up date used for retention. A temporary job, an unrelated income job, or a paid placement can stabilize income without being the main occupational outcome. A course referral becomes a transition pathway only when it connects to a destination employer or verified vacancy source.
Why is a graduate's new wage not enough to show that training paid off?
A post-programme wage is a level, not a return. It needs comparison with the worker's pre-separation wage, hours, foregone earnings, tuition, direct expenses, and the realistic wage that might have been earned without the programme. Where possible, it also needs a comparison group. Per Scholas follow-up earnings exceeded controls by $5,186 after quarter 4; that finding is stronger than a graduate-wage advertisement, but it is not a universal payoff guarantee.
Why should a provider name employers before taking a deposit?
A named employer list makes the claimed occupational destination inspectable. It permits checks of current vacancies, entry requirements, wages, schedules, geography, and whether several employers exist. A route supported by one hiring employer or one public contract can disappear before a cohort finishes. Employer talks and advisory-board participation do not substitute for a specified hiring arrangement.
What does a free retraining course still cost?
Free tuition can leave the participant paying examination fees, equipment, travel, care, data access, and income lost while studying or waiting for a licence. OECD experimental estimates put opportunity cost at about 75% of total transition cost across occupations and scenarios. The estimate is based on origin-occupation median wages and training duration, so it identifies a cost mechanism rather than a bill sent to every participant.
How long should a provider follow graduates after placement?
Follow-up should continue beyond the first job offer and record employment, occupation, employer, hours, pay, and retention at stated dates. Reskilling guidance identifies the first 30 to 180 days after placement as a period when practical barriers can end a new job. Year Up employment at four years was 82% for programme members and 80% for controls, showing why employment alone can miss job-quality and earnings differences.
What do WorkAdvance costs say about the price of a good programme?
They do not set a tuition price. In the four-site WorkAdvance evaluation, in-program operating spending ranged from $5,200 to $6,700 per participant, while three sites ranged from $5,200 to $5,800. The figures include a pathway with coaching and employer services, so they show that delivery costs extend beyond classroom instruction. They do not prove that another provider offers comparable value.
What is the difference between gross cost and net cost in retraining?
Gross cost is the spending needed to operate the programme. Net cost subtracts services participants would have used without it, such as other training or employment services. WorkAdvance estimated net costs of about $3,500 per participant at Per Scholas and $4,900 to $5,900 at the other sites. Neither figure includes every participant burden or proves a benefit, because payer and counterfactual remain separate questions.
Does a short course always make financial sense because it gets people working sooner?
No. Shorter training can reduce time away from earnings, but the destination may offer lower pay or unstable work. OECD modelled average total transition costs of about USD 7,000 for routes up to six months, USD 13,000 for six months to one year, and USD 50,000 for one to three years. Those are PPP-based experimental averages across countries, not a tuition menu or a promise about one worker's return.
Can a strong programme evaluation guarantee entry for every applicant?
No. An evaluation describes results for a studied group under stated conditions. It cannot guarantee an opening for older workers, people unable to absorb unpaid practice or lost income, or entrants competing in a crowded destination without its network. Access conditions affect both who can participate and whose outcomes appear in the data, so an average result can conceal a route that is not feasible for a particular applicant.
How should a worker use a four-question retraining test?
The test asks for a completion denominator, an employment definition and timing, post-programme wages against pre-programme wages, and named destination employers. One point is awarded for each answer independently documented by a source other than a sales claim. Four points supports proceeding, two or three points requires investigation, and zero or one point supports withholding payment until the missing evidence appears.