A group medical premium is not a mystery — it is a base rate, adjusted for who is in the group, what they claimed, and what medical care costs this year. Here is the arithmetic, step by step.
A group medical premium looks like a single opaque number, but it is assembled from parts that any buyer — or any platform embedding medical cover — can understand. The short version: the census tells the insurer who is being covered, rate tables convert that into a base price, loadings adjust for group-specific risk, and the group's own claims history rewrites the number every year at renewal. Understanding each step is the difference between negotiating a renewal and merely receiving one.
Step one: the census sets the base
Everything starts with the member list. Age and sex are the dominant rating factors: medical claims cost follows a rough U-shape over life, high in early childhood, low through the twenties, then climbing steadily. Maternity-age members carry a predictable cost that rate tables price explicitly. Dependents matter as much as employees — a group of 50 staff with 150 dependents is a 200-life risk, and a census that undercounts dependents produces a quote the insurer will correct later, on worse terms.
Insurers hold rate tables — a price per life per age band per benefit tier — and the first pass of any group quote is simply the census multiplied through those tables. This is why census accuracy is not an administrative nicety. It is the price.
Step two: loadings and discounts
The table price assumes an average group. Loadings adjust for the ways this group is not average.
- Industry loading: some occupations carry higher expected claims, and insurers price sectors accordingly.
- Network and benefit loading: a wider hospital network, lower deductibles, richer maternity or dental limits — each expands expected claims and is priced in.
- Group size adjustment: small groups get less individual treatment because one bad year in a ten-life group is statistical noise, so insurers pool them; large groups earn experience-based pricing, for better or worse.
- Regulatory minimums: in mandated markets, the benefit floor is fixed by the regulator, which compresses how much of the price benefit design can move.
Discounts run the same logic in reverse — a young workforce, a lean network, high deductibles where permitted. What buyers often miss is that loadings are an argument, not a verdict: an insurer's assumption about your group can be challenged with better data.
Step three: the loss ratio does the talking
Once a group has a year of history, its loss ratio — claims paid divided by premium earned — becomes the loudest voice in the room. An insurer that paid out most of what it collected, after its own expenses, made little or nothing on the account. The renewal quote is where it says so.
The mechanics are fairly standard. The insurer projects last year's claims forward, adds medical inflation (the annual rise in the cost and use of care, which runs well above general inflation in most markets), adjusts for any census changes, and adds expense and margin. If last year's loss ratio was low, some of the benefit may be passed back; if it was high, the increase can be steep. Groups are often surprised that a quiet year still brings an increase — that is medical inflation doing its work independent of the group's own behaviour.
The renewal dance
Renewal negotiation has a rhythm experienced buyers recognise.
- The insurer opens with a projected increase built on conservative assumptions.
- The buyer or broker counters with the census changes the projection missed — leavers, a younger intake, a large one-off claim that will not recur.
- Benefit design gives both sides room: a deductible change or network adjustment can absorb part of an increase without a headline benefit cut.
- Market testing disciplines the incumbent — a credible alternative quote moves numbers more than any argument.
- One-off large claims deserve scrutiny: pricing next year as if a rare event will repeat annually is the most common form of renewal over-projection.
The dance has a limit worth being honest about. A group whose members genuinely claim more each year will pay more each year; negotiation reallocates the increase, it does not repeal it. And switching insurer to dodge a claims-driven increase works at most once — the new insurer sees the same history, and continuity of care for members mid-treatment has real costs a spreadsheet does not show.
What this means for embedded distribution
For platforms embedding group medical — HR systems quoting at onboarding, for instance — the pricing mechanics dictate the product mechanics. Quoting requires a clean census, so census capture is the core engineering problem. Renewals arrive annually with movement in the price, so the renewal journey matters more than the first sale. And because the loss ratio drives everything, the platforms that help employers manage claims experience — accurate member data, prompt deletions of leavers, sensible network choices — are quietly helping manage the premium too. The arithmetic is the insurer's, but most of its inputs belong to whoever holds the data.