Groceries: methods

Pre-registered July 2026, before any store exists. This page states how the city-owned grocery promise will be scored, so the methodology can't bend to the results.

The promise

Five city-owned grocery stores, one in each borough, aimed at keeping food prices down. Campaign estimate was $60 million for the pilot; the administration's current estimate is about $70 million. Two stores have dates: Hunts Point in the South Bronx (about $10 million, scheduled 2027) and La Marqueta in East Harlem (about $30 million, scheduled 2029). Brooklyn, Queens, and Staten Island are still in site search.

Metric 1: the count, against the dates

Each borough's store moves through three stages: site named, construction or fit-out underway, doors open. Stores with announced dates are scored against those dates - on track, behind, or met. Boroughs without a site yet sit at announced. The promise as a whole is met when five stores are open. No citywide deadline has been announced; if one is, it gets added here with a date and a source, not retroactively judged.

Metric 2: the prices, once doors open

The stated purpose is cheaper groceries, so that is what gets measured. Starting the first full month a store is open: a fixed basket of staple items, priced on the same day at the city store and at two or three nearest full-service supermarkets within a mile. Same items, same sizes, store brands compared to store brands. The number reported is the basket gap - what the city store's basket costs relative to the neighborhood's, in percent.

The basket's exact contents will be published on this page before the first store opens, and then frozen. Receipts are photographed and kept. If an item disappears from shelves, its substitution rule is documented here before the swap counts.

Data

• Store milestones: city announcements, EDC and Council budget documents, each cited on this page as it lands.
• Prices: in-person fieldwork by this site, monthly, receipts kept. There is no public dataset of shelf prices; that is why this measure does its own collecting.

What could go wrong with this method

• Fieldwork is one person's, monthly, self-funded. A missed month will say so on the page rather than interpolate.
• A basket is a sample, not a cost-of-living index. The gap number describes those items in those stores, nothing wider.
• Store dates may move. The record here keeps both the original and the revised date, with sources.