Re-Reading My Own Refund Piece: The Split Matters More Than the Triage

What I wrote

The refund piece (published April 17) framed the story as bigger refunds, households triaging, not spending. I quoted BofA Institute's David Tinsley on the 20% debt-paydown bump in the three weeks post-refund and cited the ~30% debt allocation among lower-income filers. That data is real. But I picked one side of Tinsley's framing.

What the full data actually says

Tinsley's own word for the broader refund-week behavior was sugar-rush, not triage. BofA Institute card data shows discretionary spending up in the weeks after refund arrival: electronics, hotels, lodging, restaurants. Separately, a BofA Global Research survey found over one-third of refund recipients plan to use the cash to reduce debt -- call it ~33%. That's a plurality, not a majority, and it coexists with visible spend-mode behavior in the same BofA data.

The accurate read: refund season is a split. Spending AND debt paydown AND savings contributions, all at once, with the mix tilted by income. NPR summary.

Where the original thesis still holds

The low-income tail is where cycle risk shows first. The bottom quintile is allocating ~30% of refunds to debt service. That's a stretch signal. The cohort that triages refunds first is the same cohort driving rising credit card charge-offs and subprime auto delinquency. That fragility doesn't show up in the median household because the median is absorbing the 10% refund bump in discretionary spend instead.

The refund bump is real stimulus. 2026 refunds are averaging $3,571 through March 20 -- up 10.9% YoY per IRS data. The driver isn't higher withholding. It's the One Big Beautiful Bill Act: deductions for tips, overtime, auto loan interest, and senior citizens. That's fiscal transfer in refund form.

Spending and triage are both up. Households aren't choosing between them -- they're doing both because the 10% bump is finally enough to absorb the triage AND fund the discretionary spend. That's a more encouraging signal than my original piece conveyed.

Where I got it wrong

Headline said paying down debt, not spending. That's the low-income tail, not the median. It should have said a split, with triage concentrated at the bottom.

The financial triage report framing oversells weakness. The more honest read is that households are less stretched than they were in 2022-2023, when refund-induced debt paydown was closer to 40% at the low end. The ratio is compressing. That's a modest positive signal under the aggregate noise.

What I'd watch next

If BofA Institute's weekly card-data updates show discretionary spend rolling over by mid-May while debt paydown share rises, the triage thesis moves back up. If both keep rising and April payrolls come in soft, that's a different and worse story: refund-funded last gasp of consumption before labor income cracks.

Through end of month:

• Next BofA Consumer Checkpoint update
• April retail sales (control group)
• Continuing jobless claims -- currently elevated
• Credit card charge-off data from bank Q1 earnings (reporting now)

The refund story is worth telling. The frame needed correction. The consumer isn't collapsing, but the bottom 20% is running tighter than the top -- and that's where cycles always break first.