The state of fertility benefits

Finding one: a mandate covers far fewer people than the map suggests

Coverage maps are the most-shared artefact in this field and the most misleading. 26 of the 51 jurisdictions we track have some fertility insurance law, and 15 states and DC require IVF coverage for at least some plans. Read as a map, that looks like it covers most people.

It is not, because a state insurance mandate generally reaches only plans that state insurance law regulates — broadly, fully insured plans. Self-funded private-employer plans, where the employer pays claims from its own funds and a carrier merely administers them, generally sit outside state insurance mandates because the federal ERISA law governs them. According to KFF, about 61% of covered workers are in self-funded plans.

The practical consequence: two people living on the same street, in a state with a strong IVF mandate, can have entirely different coverage because one works for a self-funded employer. The mandate map tells you about a legal regime, not about a person.

Finding two: “must offer” is not “must cover”

One state (Texas) falls into a category most summaries collapse into the mandate count: the law requires that IVF coverage be made available to employers, not that plans include it. An employer can comply with such a law and buy nothing.

We keep this category separate because collapsing it inflates the headline number and sets up patients for an unpleasant phone call. It is the clearest example of why the useful question is never “does my state mandate IVF?” but “is my plan fully insured or self-funded, and what did my employer actually buy?”

Finding three: the employer is the real policymaker

Put findings one and two together and the conclusion is uncomfortable for advocacy and useful for employees: for the roughly 6 in 10 covered workers in self-funded plans, state mandates generally do not apply, so the entity deciding whether they can afford IVF is their employer, not their legislature. In the 25 states with no law at all, that is true by definition.

This is why we publish an employer cost model publicly rather than behind a form. A benefits team can change a plan design in one cycle; a statute takes years. The lever most people can actually reach is the one inside their own company.

Finding four: coverage is not access

The least-discussed failure mode is a benefit that is paid for and cannot be used. An IVF cycle requires roughly six to ten monitoring visits — ultrasound and bloodwork — usually early in the morning, often scheduled only the day before, and almost always at the treatment centre.

For a member two hours away, or on rotating shifts, that schedule can be the binding constraint, not the money. Plans and funds measure network adequacy by provider counts, which does not capture whether someone can attend a 7am scan every other day for a fortnight. A covered cycle abandoned halfway produces cost with no outcome, and it does not show up as a coverage problem in any report.

Finding five: the definition excludes people the plan meant to cover

Many plan documents still define infertility by a period of unprotected heterosexual intercourse. Written that way, the definition cannot be satisfied by a single person or a same-sex couple regardless of medical need, so coverage never triggers. Employers who believe they offer an inclusive benefit frequently have this language sitting in their plan document unread. It is a drafting problem with a drafting fix.