Follow-up from #8086 (17-state CHIP premium framework) and #8090 (SPM integration).
Problem
NY Child Health Plus premium schedule currently encoded (ny/hhs/chip/premium/per_child.yaml) charges:
| FPL band |
Per child / mo |
Family cap |
| 0 - 222% |
$0 |
$0 |
| 222 - 250% |
$15 |
$45 |
| 250 - 300% |
$30 |
$90 |
| 300 - 350% |
$45 |
$135 |
| 350 - 400% |
$60 |
$180 |
| >400% |
$0 (not encoded) |
$0 |
Per NY DOH, families above 400% FPL pay the full unsubsidized plan premium, which varies by health plan. We currently return $0 for these households, which understates out-of-pocket cost for higher-income CHIP-enrollees.
Same pattern as PA CHIP (skipped entirely in #8086 because CHIP premium varies by MCO from ~$45-$380/month across plans and tiers).
Why this matters
The families affected are a small share of CHIP enrollment nationally, but the dollar magnitude per household is large (hundreds of dollars per month of unmodeled cost). Cliff visualization will show a sudden $0 premium at 400% FPL where reality is a spike.
Options
- Approximate with a representative unsubsidized benchmark (statewide average of NY CHP plan premiums). Still per-child with the family cap pattern. Document as an approximation.
- Skip the >400% FPL tier (current state) but add a
null / NaN return that surfaces as "not modeled" in downstream calculators rather than $0.
- Model by MCO. Requires a
chip_plan input variable, substantial encoding effort, and is likely out of scope for a microsim that doesn't have plan-choice simulation.
Recommend (1) with a representative rate until the broader plan-choice question is addressed.
Related
Follow-up from #8086 (17-state CHIP premium framework) and #8090 (SPM integration).
Problem
NY Child Health Plus premium schedule currently encoded (
ny/hhs/chip/premium/per_child.yaml) charges:Per NY DOH, families above 400% FPL pay the full unsubsidized plan premium, which varies by health plan. We currently return $0 for these households, which understates out-of-pocket cost for higher-income CHIP-enrollees.
Same pattern as PA CHIP (skipped entirely in #8086 because CHIP premium varies by MCO from ~$45-$380/month across plans and tiers).
Why this matters
The families affected are a small share of CHIP enrollment nationally, but the dollar magnitude per household is large (hundreds of dollars per month of unmodeled cost). Cliff visualization will show a sudden $0 premium at 400% FPL where reality is a spike.
Options
null/NaNreturn that surfaces as "not modeled" in downstream calculators rather than $0.chip_planinput variable, substantial encoding effort, and is likely out of scope for a microsim that doesn't have plan-choice simulation.Recommend (1) with a representative rate until the broader plan-choice question is addressed.
Related