Regulators Approved 66 Percent of Requested Rate-Case Dollars in 2025, and Minnesota Must Decide by December 15 Whether to Move New Large Loads Out of the Commercial Class

State utility commissions granted 66 percent of the dollar value that utilities requested in 2025. Across the previous two decades the average was 52 percent. Commissions rejected only 2 of 83 requests outright.

The figures come from an analysis circulated to commercial energy buyers on September 8 and reported by Food Industry Executive. The 52 percent benchmark describes a regulatory posture that no longer holds.

The approval share.

Fourteen points of difference between the historical average and the 2025 result is a change in what a filing predicts, not a change in what utilities ask for. A buyer working from the two-decade average treats a filed increase as roughly a coin flip. At 66 percent, a filed increase is a much better forecast of the tariff that eventually bills.

The volume behind those filings is not small. Charles Hua, founder of the advocacy organisation PowerLines, told Latitude Media the same week that utilities have queued roughly $1.4 trillion of capital expenditure through 2030, and argued that the drivers of rising bills predate the data-center boom.

The class order inverted in 2025.

From 2024 to 2025, industrial retail prices rose 6.0 percent, commercial 5.2 percent and residential 5.0 percent. Industrial prices are up 27 percent since 2019.

The industrial class leading a single year of increases is a departure from the pattern commercial energy models generally assume, which is that large-volume customers absorb price pressure more slowly than households do. One year does not establish a trend. It does establish that the ordering is not fixed.

Minnesota’s deadline.

Minnesota Statute 216B.1622 requires the state commission to establish by order, no later than December 15, 2026, the definition and appropriate characteristics of a very large customer class or subclass for each public utility providing electric service. The statute sets the deadline and the deliverable. It does not settle where the thresholds land, and the commission has not yet issued the orders that will.

The question in front of the commission is a cost-allocation question rather than a siting or generation one. Carving the largest loads into a class of their own re-allocates fixed costs across the classes that remain. Whether that arithmetic moves demand-related cost recovery up or down for a mid-size commercial customer, an office or healthcare building in the 200 kW to 2 MW range, depends entirely on how the order is written. The decision is three months out, and it applies to Xcel Energy and Minnesota Power alike.

The national count.

The Edison Electric Institute counted 24 states with at least one approved large-load tariff as of July 2026, with six more pending before commissions. Minnesota’s proceeding is therefore not an outlier. It is a late entry in a category that already covers roughly half the country, and it carries a statutory deadline that most of the others did not.

What the carve-out removes.

A large-load class does two things at once, and commercial customers tend to notice only the first.

The first is protection. Transmission and distribution costs incurred to serve a new very large customer can be assigned to that customer rather than spread across the general commercial class. A commercial building does not pay for a data center’s substation.

The second is subtraction. Load growth inside a rate class spreads fixed costs over more kilowatt-hours, which is a downward force on the price every member of that class pays. Directing the next increment of large load into a standalone class removes that force from the classes left behind. The residual class then faces the revenue requirement on its own volume growth.

Both effects are real and they run in opposite directions. Which one dominates for a given customer is an empirical question about the specific order, not a matter of principle, and it is the question worth putting to each of the 24 states plus Minnesota rather than answering once nationally.

The instrument problem.

Retail price indices report total customer bills, including volumetric, demand and fixed charges, divided by total retail sales. A 5.2 percent commercial increase therefore carries no information about how much of that increase landed on billed demand rather than on energy.

Commercial storage economics turn almost entirely on the demand component. No widely cited national series reports that component separately. The class-average index that regulators, analysts and buyers all reference is consequently the wrong instrument for the question storage payback depends on, and using it requires an assumption about the split that nobody publishes.

What follows.

Two changes are underway at once. Filed rate increases are surviving regulatory review at a materially higher rate than the twenty-year average, which raises the expected value of any pending filing in a buyer’s territory. And a growing number of states, Minnesota by statute before December 15, are deciding whether the next increment of large load sits inside the commercial class or beside it.

Commercial energy models that carry a flat escalator, or one drawn from recent class-average increases, embed the historical approval haircut and the historical class structure simultaneously. The first of those inputs is stale by fourteen points. The second has a date attached to it.

Sources