Charted: Gig-Primary Households Show a 2.8 pp Card Delinquency Gap vs Matched W-2 Peers
Desk-harmonized CFPB credit-panel and Fed SHED cuts put gig-primary card 60+ DPD at 8.4% versus 5.6% for matched W-2-only households — a 2.8 percentage-point gap — with a 19 pp thinner $400 cash buffer.
Loading interactive charts…
Platform work is no longer a footnote in household cash flow. Federal Reserve Survey of Household Economics and Decisionmaking (SHED) modules put any gig or side-platform participation near 16% of adults, while a thinner slice — about 4% in our desk cut — reports platform earnings as half or more of household income. That gig-primary group is small enough to miss in headline delinquency charts and large enough to matter for revolving-credit stress, because income arrives in irregular pulses rather than a predictable paycheck.
The desk question for this note is narrow: do households that rely on platform income show different delinquency and cash-buffer patterns than comparable W-2-only households? Comparable means matched on broad income band and age, not a raw population average that mixes retirees, dual-earner professionals, and gig drivers into one mean. The interactive dashboard above synthesizes CFPB Consumer Credit Panel–style past-due rates with SHED emergency-expense buffers and a volatility map that links month-to-month income swing to card 60+ day past-due (DPD) rates.
Headline: a 2.8 percentage-point card delinquency gap
On revolving credit, gig-primary households print an estimated 8.4% card 60+ DPD rate against 5.6% for matched W-2-only peers — a 2.8 percentage-point gap. That is the headline number. Auto loans show a 2.1 pp gap; unsecured personal loans a 3.1 pp gap; first-lien mortgages stay under 1 pp. Secured credit compresses the differential; revolving and personal credit amplify it.
The gap is not a pandemic artifact. Desk vintages from 2019 through 2024 show the card differential compressed to roughly 1.3–1.4 pp in 2020–21, when stimulus and forbearance lowered past-due rates across segments, then widened every year through 2024 as revolving stress returned. Gig-primary delinquency climbed faster than the W-2 path once the temporary supports faded. For credit-risk desks that only watch the aggregate card 60+ series, that separation is invisible — the national print averages away a dependence premium that is economically large for the households that carry it.
How we define gig-primary, gig-secondary, and W-2-only
Definitions matter more than branding. Gig-primary means platform or contingent platform-like earnings account for ≥50% of household earnings in the reference window. Gig-secondary means some platform earnings under that threshold — typically a side stream on top of W-2 wages. W-2-only means no reported platform earnings in the matching frame, drawn from the same income and age cells so the comparison is not "gig workers versus everyone else."
SHED does not publish an official three-way CCP delinquency table. The rates here are desk-harmonized: CCP-style product past-due rates are applied to segment cells whose income and age mix approximate SHED gig participants, with estimated labels where the join is imperfect. Treat the levels as research synthesis, not a regulatory tabulation.
Population shares in the dashboard — roughly 4% gig-primary, 12% gig-secondary, and a large W-2-only majority among working-age adults — are order-of-magnitude guides. Exact SHED wording on "gig," "side activity," and "app-based work" shifts across vintages; we keep the labels stable for charting. Secondary earners are the majority of platform participants; primary-dependent households are the minority that drives most of the credit differential.
Product gaps: revolving and personal credit lead
Toggle Product gaps in the dashboard. Grouped bars show 60+ DPD by product for all three segments. Three patterns stand out:
- Cards dominate the absolute gap. Gig-primary at 8.4% versus W-2 at 5.6% is the clearest revolving stress signal. Utilization also runs hotterabout 62% mean revolving utilization for gig-primary versus 44% for matched W-2 — so the same income miss hits closer to the limit.
- Personal loans show the widest relative gap (+3.1 pp). That product often fills short-term cash holes when platform weeks go soft; higher past-due rates are consistent with that use case.
- Mortgages barely move. A 0.9 pp gap on first liens is economically small and may partly reflect selection: households that clear underwriting already look different on credit history, even when current income is volatile.
Filter by product to isolate one bar cluster. The story does not reverse: every product shows gig-primary above W-2; only the magnitude changes. Auto sits between cards and mortgages, as expected for a secured installment product with less discretionary drawdown than revolving lines.
Cash buffers: a 19-point $400-expense gap
Delinquency is a lagging credit outcome. Liquidity is the leading one. On the SHED-aligned ability to cover a $400 unexpected expense with cash or cash-equivalent, gig-primary households sit near 48%, gig-secondary near 58%, and matched W-2 near 67% — a 19 percentage-point buffer gap between gig-primary and W-2.
The Cash buffers panel traces 2019–2024. Every segment improved in 2020–21. Gig-primary peaked near 54% able to cover $400, then slipped back toward the high-40s as inflation and the end of transfers hit irregular earners first. W-2 buffers settled near two-thirds and held. The parallel would borrow or sell coping share runs the other way: 41% gig-primary versus 24% W-2, consistent with thinner liquid reserves.
Buffers and delinquency are not the same population event — a household can miss a card payment without failing a hypothetical $400 question, and vice versa — but they travel together in the segment averages. Thin buffers make a soft platform week more likely to become a past-due cycle. That is the operational link between SHED liquidity modules and CCP past-due tapes: cash-flow irregularity shows up first as buffer failure, then as missed revolving minimums.
Volatility map: swing predicts past-due inside income bands
Income level does not exhaust the story. Month-to-month swing matters. Desk medians put gig-primary MoM income variation near 28%, gig-secondary near 14%, and W-2 near 6%. The Volatility map scatter plots those swings against card 60+ DPD inside income bands (<$40k, $40–80k, >$80k).
Within every income band, gig-primary cells sit further right (more swing) and usually higher (more past-due) than W-2 cells. Low-income gig-primary is the stressed corner — roughly 11% card 60+ at a 34% median swing. High-income W-2 is the calm corner — about 3% past-due at a 4% swing. Gig-secondary occupies the middle of the cloud.
Filter the scatter by segment to see each cloud alone. The slope is not a causal claim: volatile work sorts into certain credit products and utilization patterns. It is a descriptive regularity that survives the income-band cut — which is the minimum bar for arguing that volatility, not only poverty, belongs in the monitoring frame.
Reliance ladder: outcomes step with platform share
If platform income is continuous rather than binary, outcomes should step as the platform share of earnings rises. The Reliance ladder does that: 0% (W-2 only), 1–24%, 25–49%, 50–74%, and 75%+. Card 60+ DPD rises from 5.6% to 9.1% across those rungs; $400 coverage falls from 67% to 44%; revolving utilization climbs from 44% to 68%.
Toggle the ladder metric among delinquency, buffer, and utilization. The monotone pattern is the finding: dependence intensity, not merely "any gig participation," tracks credit and liquidity stress. That is why blending all platform participants into one "gig worker" average flattens the risk picture — secondary earners dilute the primary-dependent tail. Lenders and policymakers who only see "gig" as a binary flag will understate stress in the ≥50% earnings group and overstate it among casual side-hustlers.
Vintage: the gap after support winds down
The Gap vintage panel overlays gig-primary and W-2 card 60+ rates with the percentage-point gap on a second axis. 2019 opens near 1.9 pp. 2020–21 compress. 2022–24 reopen to 2.8 pp. The path matches a simple narrative: temporary income support and forbearance reduced missed payments broadly; when those faded, households with choppier cash flow re-separated from paycheck peers.
This is not a claim that platforms caused the 2023–24 revolving cycle. Macro rates, utilization, and lender risk appetite moved for everyone. The claim is thinner and more useful: among households with similar income and age, platform dependence correlates with a durable delinquency premium that widened as the support overlay lifted. Watching only the national card past-due series will keep missing that wedge.
Outcome table and reading rules
The Outcome table panel (and the markdown table below) consolidates the desk metrics. Use it as a one-screen brief; use the charts for composition and vintage.
| Metric | Gig-primary | Gig-secondary | W-2 only | Note |
|---|---|---|---|---|
| Card 60+ DPD rate | 8.4% | 6.5% | 5.6% | Matched income/age CCP cells |
| Auto 60+ DPD rate | 5.9% | 4.4% | 3.8% | Matched income/age CCP cells |
| Revolving utilization | 62% | 51% | 44% | Mean utilized / limit |
| Can cover $400 expense | 48% | 58% | 67% | SHED cash/equivalent |
| Would borrow or sell | 41% | 32% | 24% | SHED emergency coping |
| Median MoM income swing | 28% | 14% | 6% | Platform vs paycheck volatility |
Caveats and what this does not show
Matching is imperfect. CCP and SHED are different samples with different gig identifiers. Desk joins on income and age reduce, but do not eliminate, selection bias — people who choose platform-primary work may differ on unobserved credit behavior.
Estimated rows are labeled. Where CFPB or Fed publications do not publish the exact crosstab, we carry forward adjacent cells or apply Chase Institute–style volatility patterns to CCP past-due rates. Do not cite a single cell as an official government statistic.
Mortgage and auto underwriting select survivors. Lower gaps on secured products partly reflect who gets the loan, not only who pays on time after origination.
Platform mix shifts. Delivery, rideshare, and task platforms have different pay volatility and tip shares; a national average hides that. Regional labor markets and app take-rates also move.
No causal identification. We do not instrument platform entry or randomize income volatility. The dashboard is a monitoring lens for delinquency and buffer gaps, not a structural model of credit risk.
Why the 2.8 pp gap is the desk number
For consumer-finance monitoring, the useful headline is not "gig workers are riskier" in the abstract. It is that households for whom platform income is the paycheck show a roughly three-point card delinquency premium versus matched W-2 peers, with a nineteen-point thinner $400 cash buffer and roughly four times the month-to-month income swing. Product gaps, the reliance ladder, and the post-stimulus vintage all point the same direction: dependence intensity and cash-flow volatility travel with revolving stress.
That framing keeps the conversation on measurable credit outcomes — past-due rates, utilization, emergency liquidity — rather than on culture-war debates about gig work itself. Platforms can be flexible income for secondary earners and a hard primary paycheck for others. The credit tape treats those groups differently. The dashboard is built so you can see that split without averaging it away.