community vs experience rating
Suppose two companies buy the same health plan. One is full of healthy 25-year-olds; the other is older and sicker. Should both pay the same premium per person, or should each pay according to its own claims? That choice — between charging everyone a shared rate and charging each group for its own experience — is the heart of community versus experience rating.
Community rating sets one premium for everyone in a defined community (or by a few broad factors like age or region), regardless of an individual's or group's own health and claims. It spreads risk widely and is fair in a social sense, but a healthy person subsidizes a sicker one. Experience rating instead bases a group's premium largely on its own past claims experience: a group that ran a 95% loss ratio pays more next year; one at 70% pays less. In practice insurers often blend the two using credibility — a large group with lots of data gets a premium mostly from its own experience, while a small group, whose few claims are too random to trust, gets a premium mostly from the broader manual (community) rate, perhaps weighted z toward its own experience and (1 - z) toward the manual.
This is everyday health-actuarial work: deciding how much to trust a group's own numbers (its credibility z) and blending experience with the manual rate. Pure experience rating is unstable for small groups (one bad year would whipsaw their premium); pure community rating invites antiselection (healthy groups leave for cheaper deals, leaving the sick behind and driving the pooled rate up). The art is choosing the rating basis and credibility that are stable, fair, and competitive at once — and the rules are often constrained by regulation.
A 2,000-employee firm has its own credible claims history, so its renewal premium is set mostly from experience: a low loss ratio earns it a discount. A 12-person firm has too little data to trust, so its premium comes mostly from the manual rate, with only a small weight z (say 0.2) on its own experience and 0.8 on the manual.
Big groups are priced on their own experience; small groups lean on the pooled manual rate.
Neither is simply 'better'. Experience rating is unstable for small groups; community rating invites antiselection unless everyone must participate. Most real systems blend them via credibility — and the allowed mix is often set by regulation, not just by the actuary.