Food Price Forecast 2027: USDA's Numbers for Restaurant Ingredients, and How Far to Trust Them

The figures below come from ERS's Food Price Outlook, published September 25, 2026 with BLS price data through August. ERS updates it about the 25th of each month; the next releases are scheduled for October 23, November 24 and December 23, 2026. Each percentage compares one year's average price with the previous year's average, so "+2.6% in 2027" means the 2027 average against the 2026 average — not a December-to-December change and not a forecast for any single month.

The 2027 forecast for menu prices and ingredients

Restaurant prices. Food away from home covers what diners pay at restaurants and other foodservice. ERS expects 3.5% in 2026 — exactly its 20-year average — and 2.6% in 2027. In August 2026 the index was 3.4% above a year earlier, against 2.2% for groceries. It is the only line in the table whose 2027 interval does not include a decline.

Beef carries the largest increase among the proteins: +11.6% in 2025, +9.4% expected in 2026 and +4.7% in 2027. For 2026, ERS points to supply — federally inspected beef production was "about 2 percent lower year over year" and is expected to stay below year-ago levels through the end of 2026. ERS publishes no explanation for any 2027 figure, and apart from the three headline aggregates, its 2027 numbers appear only in its data tables. For what beef costs now, see the BLS series on the ground beef price page.

Eggs fall 29.4% in 2026 after rising 21.9% in 2025, as laying flocks recovered from avian influenza. The 2027 midpoint, +0.8%, sits inside an interval from −30.3% to +47.8%, which says little either way. Coffee has no forecast line of its own: ERS reports retail coffee 6.1% above a year earlier in August 2026, faster than other nonalcoholic beverages, and gives no cause.

Wholesale and farm prices: closer to your invoice, harder to forecast

A restaurant buys further up the supply chain than a shopper, so ERS's Producer Price Index forecasts are the nearer signal. ERS describes the relationship plainly: PPIs "are typically far more volatile than the CPIs", and "the CPI typically lags movements in the PPI." A PPI measures prices received by domestic producers. Your distributor's price adds processing, freight and margin, and it can move later or by a different amount.

Two readings need care. Wholesale beef is forecast flat for 2027 (−0.4%) while retail beef is forecast up 4.7%. The two come from separate statistical models, each fitted to its own price history, and ERS's methods report found that adding wholesale prices to its retail pork model made the forecast worse, not better — so the retail and wholesale numbers are not expected to agree. And farm-level eggs: +21.0% in 2027 follows a forecast fall of 79.2% in 2026, so it is a percentage of a collapsed base, with an interval running to +336%.

Wholesale fats and oils are the line a fryer-heavy kitchen should watch: +19.2% expected in 2026 and +6.9% in 2027, while the retail fats and oils forecast is −0.8% for 2026. ERS publishes no explanation for either number.

How far to trust a forecast 16 months out

ERS is explicit about what the numbers are: "Forecasting methods are based entirely on statistical models that are fitted to recent trends in the data." The models extend each price series' own recent behavior; a drought, a tariff or a disease outbreak enters only once it shows up in prices. In ERS's own test of the method over 2011–2020, "more than 85 percent" of actual retail price changes landed inside the 95% interval — fewer than the 95% the interval implies — and for producer prices over 2014–2020 the share was 68% (ERS technical bulletin TB-1957).

We checked the years since with ERS's own files. ERS publishes the forecast each month would have produced for each year, from the data available at the time; for years before 2023 it recreated those forecasts with its current method, so they show how this method would have done rather than what ERS published then. For each target year from 2015 to 2025, we compared the forecast made in July of the year before — about the same distance ahead as today's 2027 numbers — with ERS's actual annual change.

Three things follow. First, the restaurant price forecast is the steady one: its July forecast missed by 1.4 points on average, and it came in below the actual in every year but 2024, when it was 0.4 points too high. Second, ingredient forecasts miss by several points even in calm years, and the largest misses run in both directions — beef fell 6.3% in 2016 against a forecast rise. Third, the forecast sharpens fast once the year starts: the forecast made in January of the same year missed food away from home by 0.4 points on average and held the actual inside its interval all 11 years. Beef's January miss was still 4.1 points.

2025 is the recent example. In July 2024 the method expected retail beef +1.4%, eggs −7.6% and restaurant prices +3.0%; the year delivered +11.6%, +21.9% and +3.8%. The 2027 numbers are moving too: between the July and September 2026 releases, the 2027 midpoint for retail fats and oils went from −8.8% to +3.4%, and farm-level milk from +10.4% to −6.7%.

Turning the forecast into a 2027 food budget

Weight each category's forecast by its share of your food purchases. The retail categories are a proxy — a restaurant buys different product forms, through a distributor — but they are the only USDA forecast that covers the whole basket. Use your own invoice totals for the shares.

Beef is 18% of this kitchen's purchases and 42.5% of its expected increase. That is the general pattern: the budget is driven by the one or two categories with both a large share and a large forecast, and those are also where the forecast is least reliable.

What to do with it this quarter

• —Set the number in January, sketch it now. The October 23, November 24 and December 23 releases will move these figures; the January forecast for food away from home has missed by 0.4 points on average since 2015.
• —Budget beef, eggs and fats on a range. Their midpoints have missed by 4 to 17 points; a single number for them is false precision.
• —Watch your own invoices first. Your distributor's price is the earliest signal you have, and it is the price you pay. The invoice price audit shows how to check that each delivery matches what you agreed, and the Food Cost Index tracks monthly BLS prices for common ingredients.
• —Plan around seasons separately. An annual forecast says nothing about the months within the year; seasonal food prices covers which ingredients follow the calendar.
• —Price from your plate cost, not from a forecast. A forecast tells you when to look again. The dish price comes from what the plate costs on your current invoices — how to raise menu prices works through that decision.

Sources: ERS Summary Findings and the CPI and PPI forecast tables, the historical forecast series and the annual percent-change files on the Food Price Outlook page, all as published September 25, 2026 and downloaded October 4, 2026; MacLachlan, Chelius and Short, Time-Series Methods for Forecasting and Modeling Uncertainty in the Food Price Outlook, ERS TB-1957 (August 2022). The accuracy table and the budget are Marji's arithmetic on those files. ERS forecasts are model output, not official USDA farm-price projections; ERS points readers to the World Agricultural Supply and Demand Estimates for those.