Why would increasing the SSI earned income exclusion leave households worse off?

[MaxGhenis]

Raising the SSI earned income exclusion from 50% to 75% (https://policyengine.org/us/population-impact?ssi_earned_income_exclusion_share=75) costs $9bn, leaves 2% better off, and 1% worse off. Those exact numbers will be off since we only pipe age and blindness SSI criteria from the CPS (not disability), but it's still puzzling why we'd have any losers from this policy. Playing around with household views, I couldn't identify a case where a household comes out behind. It strictly increases SSI, so when would more SSI leave a household worse off?

[MaxGhenis]

Now it leaves nobody worse off, so might have gotten fixed by a recent disability-related enhancement. Closing.