12.2 Group Underwriting and Contribution/Participation

Group underwriting evaluates the group as a whole rather than each member individually. The insurer looks at the size of the group, the nature of the business, the age and gender distribution, the geographic location, and the historical claims of the group. Because individuals are not medically underwritten, the insurer relies on contribution and participation rules to prevent adverse selection.

Think of group underwriting as answering three questions: Is this a real, stable group? Will enough people enroll to make the risk pool predictable? And how should we price the contract given the group's size and claims history? The first question is the eligibility test from the prior section; this section answers the second and third.

Contributory vs. Noncontributory Plans

• Contributory plan – employees pay part of the premium. To guard against only sick employees enrolling, insurers require a minimum 75% participation of eligible employees.
• Noncontributory plan – the employer pays the entire premium. Because the employee pays nothing, insurers require 100% participation of eligible employees.

Participation Requirements at a Glance

The purpose of both rules is the same: force a broad cross-section of the workforce into the plan so the healthy subsidize the unhealthy. If participation falls below the threshold, the insurer may decline or re-rate the group.

Trap: the percentages are reversed from what intuition suggests. Noncontributory (free to the employee) demands the higher 100% figure precisely because there is no cost barrier to enrolling everyone.

Rating Methods

The insurer sets the group premium using one of three approaches that depend largely on group size:

• Community rating – everyone in a geographic area or class pays the same rate regardless of the specific group's claims. Common for small groups.
• Manual (class) rating – rates come from the insurer's tables based on broad factors (industry, age bands). Used when a group lacks credible experience.
• Experience rating – premiums reflect the group's own claims history. Used for large groups because their data is statistically credible.

The larger the group, the more credible its experience, so large groups move toward full experience rating and small groups toward community or manual rating.

A related concept is the loss ratio — incurred claims divided by earned premium. If a group of fifty employees pays $600,000 in annual premium and generates $480,000 in claims, its loss ratio is 80%. Insurers compare this to a target (say 75–85%). A persistently high loss ratio signals that the renewal premium must rise or benefits must be trimmed, while a low loss ratio gives the broker leverage to negotiate a better renewal.

Worked Example: Participation Test

An employer has 80 eligible employees and offers a contributory plan. The insurer requires 75% participation.

• Required enrollees = 80 × 0.75 = 60 employees.
• If only 54 enroll (67.5%), the group fails the test and the insurer can decline or load the rate.

Now switch to a noncontributory plan: the employer must enroll all 80 eligible employees (100%). A single eligible employee who waives in writing can jeopardize the noncontributory status, which is why employers usually auto-enroll everyone.

Experience Rating Mechanics (Worked)

Experience-rated premiums often combine the group's actual loss ratio with the insurer's expected (manual) rate using a credibility factor (Z):

Adjusted rate = Z × (group experience rate) + (1 − Z) × (manual rate)

Example: a large group is assigned Z = 0.80. Its experience rate is $520/employee and the manual rate is $600.

• Adjusted = 0.80 × $520 + 0.20 × $600 = $416 + $120 = $536/employee.

A favorable claims year lowers next year's rate; a bad year raises it. Small groups get a low Z (little credibility), so their rates stay near the manual/community figure.

Understand why credibility scales with size. A group of ten can have one catastrophic claim that doubles its loss ratio purely by chance — that is statistical noise, not a true reflection of risk. A group of five thousand has enough claims that random spikes average out, so its experience genuinely predicts future cost. Blending with the manual rate using Z protects small employers from being punished for bad luck while still letting large employers earn lower rates for genuinely good experience.