Engine evidence · 2026 filed plans · methodology below
The cost of picking a health plan by premium.
Every plan-shopping tool sorts by premium. We simulated the whole year instead — every filed plan, every Georgia county, nine care profiles, three ages — and measured what the cheapest-sticker habit costs. The answer: the lowest-premium plan is the wrong pick 57% of the time, at a median $548 a year.
Georgia, 2026: 4,293 combinations, zero skipped.
Annual cost of the lowest-premium pick versus the lowest simulated 12-month total, across all 159 Georgia counties × 9 illustrative care profiles × ages 28/40/55.
$548/yr
Median cost of the premium pick
$2,492/yr
90th percentile
$5,489/yr
Largest observed
57% of cells
Premium pick was the wrong plan
Texas replicates the shape across 254 counties: median $435/yr, $2,381 at the 90th percentile, wrong pick 59% of the time.
When the cheapest plan is right — and when it isn't.
The honest shape of the data: for a routine, low-utilization year the premium pick is essentially always right. The cost concentrates where care is regular — and that's exactly who premium-sorting tools fail. Median annual cost of the premium pick by care profile, Georgia:
Routine year
$0
Hypertension, managed
$0
Type 2 diabetes, managed
$0
Back pain — physical therapy
$0
Regular therapy
$1,048/yr
Knee injury + rehab
$1,108/yr
Asthma — maintenance inhaler
$1,240/yr
Chronic migraine
$1,532/yr
Autoimmune — specialty biologic
$2,459/yr
Age cuts the other way from intuition: the median cost of the premium pick was $1,045/yr at age 28, $730 at 40, and $0 at 55 — the federal 3:1 age curve widens the dollar spread between plans for younger shoppers.
How these numbers are computed.
Real filed data
Premiums are real, unsubsidized 2026 individual-market rates from government filings (every premium traces to a public rate file); deductibles and out-of-pocket maxima are each plan's real filed accumulators. Snapshot snapshot_8827361338bd…, source as of 2026-06-03.
One labeled care rule — never guessed copays
Assumed care flows through each plan's real deductible and out-of-pocket maximum under one uniform rule: 100% of allowed cost until the deductible is met, then a flat 25% coinsurance, capped at the plan's real out-of-pocket max. Plan-specific copays are never guessed, so plan differences come only from real filed premiums, deductibles, and OOP maxima. Drug fills are modeled by tier at labeled assumed amounts.
Fail-closed, deterministic, replayable
Plans with missing or unverified accumulator data are excluded and counted, never assumed. Every eligible plan in every county was simulatable: 4,293 Georgia cells computed, 0 skipped. The computation is deterministic — identical reruns produce identical output — and ties at the lowest premium are resolved in the premium-sorter's favor, so the reported deltas are conservative.
What this is not
These are modeled illustrations under the labeled assumptions above — not quotes, not predictions of any individual's spending, and not advice. A person's actual best plan depends on their actual care, drugs, and doctors, which is what the full engine prices per person.
The full pricing methodology — sources, checksums, fail-closed rules — is public.
The pattern is old. The measurement is new.
Medicare Part D (Abaluck & Gruber, American Economic Review 2011)
The canonical study: typical enrollees could save roughly 30% of total costs (premiums plus out-of-pocket) with a better plan choice; only about 15% chose the lowest-cost option — and follow-up work found no learning over time.
ACA marketplace (federal ASPE analyses, 2015–2016 plan years)
Returning consumers who switched plans within the same metal level saved an average of about $400 a year (2015); the average returning silver enrollee who shopped for 2016 could save about $624. Dated figures — the mechanism, not the magnitudes, is the point.
The cheapest plan moves (KFF, 2015→2016)
In 73% of counties, the lowest-cost silver plan one year was no longer the lowest-cost silver the next — so even a correct premium pick decays. A true-cost ranking has to be recomputed every year, per person.
Plan-choice questions, answered with the data.
Is the cheapest health insurance plan ever the right choice?
Often — for a low-utilization year. In our 2026 Georgia evidence, the lowest-premium plan was also the lowest true-cost plan in 43% of county-profile-age combinations, and for a routine year with no regular care it was essentially always right. The mistake is using premium as the rule when you have a managed condition: that's where the median cost of the premium pick runs $1,240–$2,459 a year over the true-cost pick.
How much does picking the wrong plan cost?
Across every Georgia county, nine illustrative care profiles, and three ages (4,293 combinations of 2026 filed plans), choosing by lowest premium instead of lowest simulated 12-month total cost a median $548 a year, $2,492 at the 90th percentile, and $5,489 at the maximum. Texas replicates the shape: median $435, with the premium pick wrong 59% of the time.
Why does the lowest-premium plan lose so often?
Because premiums and cost-sharing move in opposite directions: the cheapest sticker is usually a high-deductible bronze design, so a year with regular prescriptions, therapy, or specialty care runs through the full deductible before the plan pays. A whole-year simulation — premium plus deductible path plus drug costs, capped at the real out-of-pocket maximum — regularly ranks a higher-premium plan cheaper by hundreds or thousands of dollars.
Who is most exposed to this mistake?
Younger shoppers, counterintuitively: at age 28 the median cost of the premium pick in Georgia was $1,045 a year, versus $0 at 55 — because the federal 3:1 age curve widens the dollar spread between plans as premiums scale, and because premium-sorting tools all present the same cheapest-first default regardless of the shopper's care.
Has this pattern been documented elsewhere?
Yes — it is one of the most replicated findings in health-insurance economics. In Medicare Part D, Abaluck and Gruber (American Economic Review, 2011) found typical enrollees could save about 30% of total costs with a better plan choice and that only about 15% chose the lowest-cost option; federal ASPE analyses of the ACA marketplace (2015–2016 plan years) found returning shoppers who switched saved roughly $400–$600 a year, and KFF found the lowest-cost silver plan changed in 73% of counties in a single year. Our contribution is current-year, county-level, filed-data evidence with a published method.
How do I check this for my own situation?
For a household, the true-cost comparison for your county, age, and care is what Insurf's engine computes. For an employer, the free audit prices your whole roster the same way. Both show the arithmetic rather than a score, and the pricing methodology — sources, fail-closed rules, what we refuse to guess — is public.
Stop sorting by premium. Price the year.
Insurf, Inc. is not a licensed insurance agency, does not sell insurance, and receives no commissions; enrollment applications are submitted by an individually licensed insurance producer. Figures on this page are modeled illustrations from filed public data under the labeled assumptions above — informational only, not an offer of coverage, not a recommendation to buy or drop any plan, and not financial, tax, legal, medical, coverage, or insurance advice. Projections are labeled and are not guarantees.