Purchasing power lab · v1
Same prices.
Different inflation.
Income before taxes, 2019–2025 · national CPI prices
Every decile below faces the same national category price changes. The bars differ only because each income group spends its budget differently. Move a category and watch who absorbs it. The purchasing-power section sets those prices against the income BLS recorded for the same decile.
01Price changes · 12-month %
02Inflation each decile actually faces
Replay · real BLS prices, Jan 2019 → Sep 2026
Each month applies that month’s 12-month change in eight BLS CPI-U series (U.S. city average, NSA) to the fixed 2025 decile baskets. October 2025 was never published (federal shutdown) and is left as a gap. Data: .
03Purchasing power · break-even and real income
The break-even column is this scenario’s own inflation rate. The real-income column stays blank until you test a hypothetical growth rate. Observed income for 2019–2025 is the next section.
| Decile | Coverage | Mapped-only | Full proxy | Gap vs CPI | Break-even wage growth | Real income change |
|---|
Subgroup overlay · renter / owner / age / region
04Observed purchasing power · calendar 2019–2025
Mean income before taxes for each decile, against that decile’s basket priced with annual-average CPI. Changing the basket vintage above changes the inflation and the real-income columns. The income dollars do not change.
| Decile | Income 2019 | Income 2025 | Nominal | Basket inflation | Real income | Dollar change |
|---|
Each year · real before-tax income on the fixed basket
05Wealth access · cost of entering home ownership
Price and rate dates are not matched (FHFA Aug→Aug; Freddie Mac Oct→Oct). This is a financing stress test, not a matched-time housing index. Excludes taxes, insurance, HOA, and closing costs.
06Independent price audit · matched observations
Same product, same size, comparable quality, same market, matched month one year apart. Rows stay in this browser tab only.
| Category | Item | Base | Current | Observed | Official proxy | Difference | |
|---|---|---|---|---|---|---|---|
| No observations yet. The model stays dependent on CPI category prices until this fills. | |||||||
What the model computes — and what it cannot see
Full-basket proxy: πg = ( Σi wg,i · πi + wg,res · πres ) / 100
Mapped-only: πg = Σi wg,i · πi / Σi wg,i
Real income: (1 + gg) / (1 + πg) − 1
Payment: L · r / (1 − (1 + r)−360), r = rate / 12, L = price · (1 − down)
- Basket composition only. Every decile is charged the national category rate. Differences in prices actually paid — store, brand, region, credit terms — are not in this model. That is the price-audit table’s job.
- Coverage differs. On the 2025 weights the seven mapped categories cover 88.4% of the bottom decile’s spending and 70.2% of the top decile’s. The remainder — insurance, pensions, education, cash contributions, apparel, and more — takes the residual rate, which is an assumption.
- Category proxies are loose. CE “Housing” is not CPI “Housing”; medical CPI is not insurance premiums; CE Transport includes vehicle purchases.
- Vintage. The default weights are the 2025 deciles. 2022, the v1 workbook, is still selectable. Prices in the scenario are Sep 2025 → Sep 2026.
- Income. Before-tax CE means, 2016–2025. After-tax income is published through 2023 and absent after that. A decile in 2025 is whoever was in it in 2025.
- What a gap measures. A gap between a decile and the headline is that decile’s basket exposure. Prices actually paid — store, brand, region, credit terms — enter only when the audit table has a matched observation.