By: Natalie Johnson
Traditional employment assumes an employee arrives stable. Housing, transportation, a bank account, and a life organized enough that the company buys only the hours. Rachel Bernier-Green has worked with a population for whom that is a distant reality. “Most of the failures that I see in the first 90 days are very rarely about the person’s capability,” she says. “They’re just a collision between the expectations of a manager or a team, and a life that is in transition and on a pathway to stability.”
Employers willing to reach past that boundary, into pay timing, transportation, and banking access, tend to see stronger retention. According to data compiled by the U.S. Chamber of Commerce, people who hold a job for a year after release return to prison at around 16% over the following three years, compared with 52% for those who cannot keep one. Retention is the intervention, built out of things that have nothing to do with the job.
Three Decisions Made Outside the Job Description
Bernier-Green identifies three decisions that can determine whether a person rebuilding a life is set up to succeed: pay cadence, scheduling predictability, and compensation.
1. Pay cadence: Someone rebuilding a life often cannot absorb a two to three week gap before a first paycheck, which turns a routine biweekly cycle into an obstacle that weekly pay can remove.
2. Scheduling predictability: Reentry means court dates, parole check-ins, treatment, and the errands required to obtain an ID or open a bank account. Nearly all of it falls during business hours. The question is how attendance and leave policies are written. A policy that codes those absences as failures penalizes the exact activities keeping someone out of prison.
3. Compensation: Research indicates that higher-paying roles reduce reoffending more than minimum-wage ones, which means the wage decision does more work than the hiring decision.
Model the Instability, Do Not Pretend It Is Free
The Federal Work Opportunity Tax Credit has historically run from roughly $2,400 to $9,600 per eligible hire depending on the target group, though its authorization lapsed at the end of 2025 and awaits congressional renewal. The federal bonding program provides free fidelity bonds covering an employee’s first six months. Bernier-Green treats both as upside, but cautions that unit economics have to work without them.
She is equally direct about the costs of absorbing instability. A slightly longer ramp to productivity, transportation support, specialized onboarding, and mentorship time. “Put a number on it instead of pretending it’s zero,” she says. What balances the equation is the return. Turnover is among the most expensive things a small business carries, and with the right supports in place, retention improves. Modeled against avoided replacement cost, the support often looks less expensive than it first appears.
What Reentry Support Looks Like in the First 90 Days
Those first 90 days carry outsized weight, and the interventions Bernier-Green describes are unglamorous. Employees need a safe route to work, since public transportation is not a safe option for everyone, which in practice meant arranging carpools, reimbursing ride-shares, or buying train tickets. They need fast and frequent pay to close the gap before the first paycheck arrives. They also may need help becoming bankable, a requirement she illustrates with employees who had the police called on them while trying to cash a check without an account.
The last support is a person rather than a process. A named mentor running structured check-ins replaces the sink-or-swim approach most new hires receive. Each of these addresses a problem that exists outside the building and shapes whether someone can stay inside it.
AI Automates the Hiring You Already Did
Roughly 88% of companies already use some form of AI for initial candidate screening, according to World Economic Forum reporting, and Bernier-Green states the mechanism plainly. A system trained on past hiring reproduces past hiring, and many companies allow it to reject candidates at any stage with no human review. Her guidance is specific: employers should audit their tools for disparate impact and ask vendors directly what the system screens on. They should remove automatic rejections triggered by employment gaps, drop check-the-box policies, and push the screening toward demonstrated skills. In some cases, she says, the better answer is abandoning the tool in favor of older hiring practices.
Prepare to Be Asked for Numbers
The goodwill era let leaders tell a story. Bernier-Green expects boards, funders, and lenders to start asking for cohort retention data, time to productivity, and avoided turnover cost. “You can’t scale or defend something that you’re not measuring,” she says.
She also flags a moving compliance floor, with more states adopting ban-the-box and clean slate laws and tax credits subject to reauthorization. Her last instruction is to build documented onboarding, manager training, and support systems, so the program outlasts whoever championed it.
None of this depends on goodwill. An employer who helps stabilize a life gets an employee who stays, which is worth more than the support costs. To learn more, connect with Rachel Bernier-Green on LinkedIn.
