Everyone agrees displaced workers need retraining. Almost nobody asks the harder question: does retraining actually work? The evidence is mixed at best โ but some program designs dramatically outperform others.
The Sobering Statistics
| Metric | National Average | Top Programs |
|---|---|---|
| Program completion rate | 15โ25% | 65โ80% |
| Employment within 6 months | 38% | 72% |
| Wage recovery (% of previous salary) | 68% | 92% |
| Still in new career after 2 years | 42% | 78% |
| Average program cost per participant | $8,200 | $14,500 |
| Cost per successful outcome | $54,000 | $19,700 |
The typical government retraining program is expensive and ineffective. But the gap between average and excellent programs is enormous โ better design, not more spending, is the key.
Why Most Programs Fail
- No income during training: Workers can't afford 6โ12 months without pay; they drop out to take any available job
- Misaligned curriculum: Programs teach skills employers don't actually need in the local market
- No employer connection: Graduates have certificates but no pathway to interviews
- One-size-fits-all: A 55-year-old bookkeeper and a 28-year-old copywriter need very different programs
- No support services: Childcare, transportation, and mental health support are absent
What Actually Works: 7 Evidence-Based Design Principles
1. Paid Training (Income Replacement)
Programs that provide 70โ80% wage replacement during training see 3x higher completion rates. Examples:
- Germany's Kurzarbeit (short-work) model: Employers reduce hours, government subsidizes wages, workers train during freed time
- Denmark's flexicurity: Generous unemployment benefits conditional on active retraining participation
- U.S. Trade Adjustment Assistance: Provides income support but only for trade-displaced workers (needs expansion to AI displacement)
2. Employer-Partnered Programs
The most effective programs have employers at the table from day one:
- Amazon's Career Choice: Pays 95% of tuition for in-demand fields; 82% placement rate
- Year Up: 6-month training + 6-month internship; 80% employment rate, $52K average starting salary
- Apprenticeship programs: 91% employment rate, $80K average annual salary for completers
3. Short, Intensive, Focused
Programs under 6 months with clear, specific outcomes outperform longer, broader programs:
| Program Length | Completion Rate | Employment Rate |
|---|---|---|
| Under 3 months | 72% | 61% |
| 3โ6 months | 58% | 68% |
| 6โ12 months | 41% | 64% |
| 12โ24 months | 28% | 71% |
| 24+ months | 19% | 74% |
Note: Longer programs have higher employment rates for completers, but much lower completion. The sweet spot is 3โ6 months for most career transitions.
4. Cohort-Based with Peer Support
Programs where participants go through as a group see 40% higher completion than self-paced online courses. Social accountability and peer support matter enormously for adult learners.
5. Wrap-Around Services
Successful programs provide: childcare assistance, transportation stipends, mental health counseling, career coaching, and interview preparation โ not just classroom instruction.
6. Industry-Recognized Credentials
Programs leading to specific, employer-recognized certifications outperform general education:
- CompTIA Security+ (cybersecurity) โ recognized by 94% of hiring managers
- AWS/Azure cloud certifications โ average $15K salary premium
- Certified Nursing Assistant โ 6-week pathway to immediate employment
- OSHA certifications โ required for many construction/safety roles
7. Local Labor Market Alignment
The best programs are designed around specific local employer needs, not generic national curricula. A program in Hartford should train for insurance-adjacent tech roles; one in Houston should focus on energy transition careers.
Top-Performing Programs in the U.S.
| Program | Focus | Completion | Placement | Avg. Salary |
|---|---|---|---|---|
| Year Up | Tech, finance | 78% | 80% | $52,000 |
| Per Scholas | IT, cybersecurity | 85% | 82% | $48,000 |
| IBEW Apprenticeship | Electrical | 75% | 91% | $62,000 |
| Climb Hire | Sales, account management | 80% | 76% | $55,000 |
| Merit America | IT, data analytics | 72% | 78% | $50,000 |
The Policy Bottom Line
We don't need more retraining programs โ we need better ones. Every dollar spent on a poorly designed program is a dollar wasted and a worker failed. Policymakers should fund programs that meet evidence-based design criteria, not just programs that exist.