TPA vs. Traditional Group Insurance: How the Two Models Actually Work

When a company buys a group health policy, it is really buying two things bundled into one price: risk transfer (the insurer promises to pay claims, whatever they turn out to be) and administration (someone enrols members, checks claims, pays hospitals and produces reports). A third-party administrator, or TPA, exists because those two jobs can be separated — and for many mid-sized employers, separating them changes the economics of the whole benefit.

What a fully-insured plan bundles together

Under a traditional fully-insured arrangement, the employer pays a fixed premium and the insurer takes on everything: actuarial risk, claims processing, provider contracts and member service. The premium is priced so that, across the insurer’s whole book of business, collected premiums exceed paid claims plus expenses. That certainty is genuinely valuable — a bad claims year is the insurer’s problem, not yours.

The trade-offs are just as structural. Premiums are recalculated every year, usually upward, and the calculation is largely opaque to the buyer. Claims data often belongs to the insurer, so an employer that wants to understand its own utilisation may receive only summary figures. And because administration is bundled, its cost and quality cannot be evaluated — or negotiated — separately.

What a TPA does — and does not do

A TPA is an administration specialist. It runs enrolment, adjudicates claims against the plan’s rules, manages the provider network and direct-billing agreements, handles pre-authorisations, and reports utilisation back to whoever carries the risk. Crucially, a TPA does not insure anyone. The risk sits elsewhere: with an insurer (which may delegate administration to the TPA), or with the employer itself in a self-funded plan.

This separation is the heart of the model. Because the administrator does not profit from underwriting, its incentives centre on processing accuracy, cost control and service quality — and its fees are visible as a line item rather than buried inside a premium.

The spectrum between the two poles

In practice, companies rarely jump from one extreme to the other. Common intermediate arrangements include:

• Fully-insured with carved-out administration — an insurer carries the risk while a TPA runs claims and the network, giving the employer better data without changing who pays claims.
• Level-funded plans — the employer pays a stable monthly amount that combines expected claims, admin fees and stop-loss insurance; surpluses may be partly returned in good years.
• Self-funding with stop-loss — the employer pays its own claims through a TPA and buys insurance only for catastrophic cases, capping the downside.

When each model tends to fit

Full insurance suits organisations that value budget certainty above all, have volatile or very small headcounts, or lack the appetite to engage with claims data. TPA-based arrangements tend to fit employers that are large enough for claims to be statistically predictable — the 100-to-500-employee range is where this usually begins — and that want transparency: to see where money goes, to keep savings from good years, and to design the plan around their own workforce rather than a standard product.

Neither model is universally better. The useful question is narrower: are you paying for risk transfer you no longer need, and would you act on the data an administrator could give you? If the answer to both is yes, the TPA model deserves a serious look.