Grocery price index

Three official US public-domain datasets, layered together: (1) BLS Average Price Data (APU) — national unit prices for ~28 staple items, published monthly by the U.S. Bureau of Labor Statistics; (2) BEA Regional Price Parities (RPPs) — state-level all-items price multipliers (US=100), published annually by the Bureau of Economic Analysis; and (3) store-tier multipliers derived from USDA ERS retail-format research and Hausman & Leibtag (2005) on supercenter vs. conventional supermarket pricing. The estimated state price is national_avg × (state_RPP / 100) × store_tier_multiplier.

Because no free, redistributable dataset of per-store grocery prices exists. Walmart, Kroger, Whole Foods and the rest don't publish their prices in machine-readable form, and scraping them is brittle and against most retailers' terms of service. The store-tier abstraction — discount (supercenters/clubs), conventional (mainstream supermarkets), premium (natural/specialty) — captures the bulk of the inter-store variation that real research has documented, while staying entirely on top of public data.

It's a model, not a price quote. BEA's all-items RPP is the standard cited figure for state-level cost-of-living differences (it's what news headlines, federal pay tables, and academic research use), but it covers all consumer expenditures, not just food. State-level food-only price parities are noisier and less consistently published, so the all-items RPP is the most defensible free state multiplier. Expect the model to be directionally correct (CA > UT, NY > MS, etc.) and within roughly ±5% of the true state average for a typical basket. Don't treat any single line item as exact.

For each historical year we scale the snapshot basket total by the ratio of that year's BLS CPI-U "Food at home" annual average (series CUUR0000SAF11) to the snapshot year's value. This captures the food-specific inflation curve — including the 2022 spike — without pretending we know the exact monthly state-level price of bananas in 2014. Years labeled estimated are partial-year extrapolations past the latest BLS publication.

The default weekly basket is every BLS APU staple, weighted by a typical weekly quantity for a 4-person household (e.g. 2 dozen eggs, 2 gal milk, 2 lb ground beef). The household size input scales every quantity uniformly. Use the Basket dropdown to switch between the full weekly mix, an "essentials" subset (bread, milk, eggs, etc.), produce-only, pantry-only, or meat & dairy.

Mostly because of the BEA Regional Price Parity (RPP). RPPs reflect retail-rent costs, wages, supply chains, and population density. California's RPP is around 113 (13% above the US average); Mississippi's is around 87 (13% below). Apply that to the same basket of national-average items and you get the spread you see. Store tier and household size scale on top of that.

Discount = supercenter / warehouse-club format (Walmart, Aldi, Costco, Sam's Club) — about 12% below the conventional baseline. Conventional = mainstream supermarket chain (Kroger, Safeway, Publix, Stop & Shop) — this is the implicit BLS APU sample mix. Premium = natural / specialty / upscale (Whole Foods, Sprouts, Erewhon) — about 15% above conventional. These multipliers come from published research on retail-format pricing, not from any single store.