Methodology

Population & unit

All SCF-based comparisons use the Survey of Consumer Finances concept of a primary economic unit (PEU) — roughly the dominant person or couple in a household plus dependents financially tied to them. We say “family” in the interface as everyday language for a PEU; it is not identical to every Census household definition.

Privacy

• Net worth and age stay on your device for the percentile math; they are not uploaded for that step.
• Optional ZIP adjustment uses only the five-digit ZIP you enter to look up published U.S. Census Bureau American Community Survey (ACS) median household income. Your net worth is not sent with that lookup.

Data & dollars

Thresholds come from the Federal Reserve SCF 2022 release (October 2023). “Current” dollars use Bureau of Labor Statistics CPI-U: 2022 annual average vs March 2026 (factor about 1.129). You can toggle to view unadjusted 2022 survey dollars.

Age bands & your exact age

The public file is grouped into age brackets (mostly five-year spans, with wider youth and open-ended upper bands). For a specific age, we linearly blend the two neighboring brackets using their midpoints so the percentile, table benchmarks, and chart all match the same blended wealth curve.

Multiple imputation

The SCF uses five imputed “copies” of the dataset. We compute weighted percentiles within each age band for each copy, then average. That yields a single combined curve — not full Rubin-style uncertainty intervals.

Optional ZIP income proxy

This is not net worth for your ZIP. We still use the national SCF curve by age. If you opt in, we compare your ZIP’s area median household income (ACS table B19013, as a ZIP tabulation area where published) to the national median, form a ratio, cap very high or low ratios so they do not dominate the chart, and multiply every dollar on the national comparison curve by that capped ratio. Your entered net worth does not change — only the yardstick does. USPS ZIPs do not always line up with a single ZCTA; we may fall back to an earlier ACS release when a row is missing.

What we do not claim

• No survey confidence intervals or full multiple-imputation variance.
• ZIP mode is not local SCF wealth; it is the national curve scaled by published median income as a rough area proxy.
• One decimal in the interface is formatting, not extra precision from the underlying sample.