The New Teacher Center, a nonprofit that has studied teacher induction and mentoring structures for two decades, has consistently found that the quality of a mentoring relationship, not merely its existence, is what predicts reduced early-career attrition — and quality depends heavily on whether a mentor has enough protected time to actually observe, plan with, and debrief a new teacher, rather than squeezing mentoring into whatever gaps remain in an already full teaching schedule. That finding is the central issue a district faces when deciding how to compensate mentors: a stipend alone, without a change to the mentor's own workload, frequently produces the second outcome.
This is a framework for that compensation decision, not a recommendation on any specific dollar figure, which depends heavily on a district's budget and local labor-cost context.
What Does a Stipend-Only Model Actually Buy?
A stipend paid on top of a mentor's full teaching load compensates the mentor's personal time — planning periods, before- or after-school hours, occasional coverage swaps — without changing the structural reality that the mentor is still teaching a full caseload of their own students. This model is administratively simple and cheaper for the district to budget, and it can work reasonably well when the stipend is paired with a genuinely light mentoring caseload, such as a single new teacher, and clear expectations about the specific time commitment involved. It tends to produce weaker mentoring, according to induction researchers, when districts assign an experienced teacher multiple mentees on top of a full course load, effectively asking the stipend to compensate for a time deficit that no reasonable stipend fully closes.
What Does Release Time Actually Require?
Release time — reducing a mentor's own teaching load, either through a reduced course section, a common planning period aligned with the mentee, or a partial reassignment to a formal instructional-coach role — directly addresses the time constraint that stipend-only models leave unresolved, and induction research generally associates it with stronger, more consistent mentoring contact. It costs more, since the district is effectively paying for a portion of a teaching position to be replaced or covered elsewhere, and it requires master-schedule flexibility that not every school, particularly a small one with limited staffing depth, can easily accommodate.
How Should a District Decide Between the Two Models?
The decision should track the actual capacity constraint a district is trying to solve. A district retaining most new teachers successfully, with a small number of struggling schools, may get more value from concentrating release-time-supported mentoring specifically at those schools rather than spreading a district-wide stipend program thin across every building. A district with a broad, moderate early-career attrition problem and limited budget for release time may reasonably choose a stipend-only model, but should pair it with an explicit, capped mentee caseload — most induction researchers suggest no more than one or two new teachers per mentor — rather than allowing caseloads to expand informally as new hires arrive.
What Should a Mentoring Program Track to Know Which Model Is Working?
Regardless of the compensation model chosen, the leading indicator to track is actual contact frequency and quality — logged observation and debrief sessions, not just the existence of an assigned mentor pairing — since induction research is specific that assigned mentoring alone, without regular substantive contact, shows little relationship to reduced attrition. A district that tracks only whether every new teacher has a named mentor, without tracking whether meaningful contact actually happened, cannot tell whether its compensation model is functioning as designed.
What Belongs in a Mentoring Program Budget Proposal?
A proposal that names the specific time-versus-money tradeoff, rather than presenting compensation as a single line item, gives a school board the information it needs to weigh the choice deliberately: the stipend cost of a stipend-only model at a capped caseload, against the staffing cost of release time at the schools with the most acute retention need, with the contact-frequency data from a prior year, where available, as evidence for which approach the district's own new teachers actually received.
