What building maintenance should actually cost.

Most numbers you will find about deferred maintenance are unsourced. Before writing this we went looking for the origin of the two most repeated ones — that every dollar deferred costs four or five later, and that reactive maintenance costs three to five times preventive. Neither could be traced to any primary study. They appear on vendor sites and blogs citing each other. So they are not in this guide.

What follows are the figures that do have a source.

The benchmark: 2 to 4 percent a year

The most durable number in this field comes from the National Research Council’s 1990 report Committing to the Cost of Ownership: routine annual maintenance and repair of a substantial building portfolio should run between 2 and 4 percent of the aggregate replacement value of those buildings. It was reiterated in the Council’s 1998 follow-up and is still the reference point in federal facilities work today.

The National Park Service states the same thing more plainly for owners of individual buildings: “Institutions generally need to budget annually between 2 and 4 percent of the replacement value of the building.”

For a building that would cost $800,000 to replace, that is roughly $16,000 to $32,000 a year across everything — roof, envelope, mechanical, finishes. Whether that sounds high depends mostly on what you have been spending.

What happens when you spend less

There is a large-scale example. The Public Buildings Reform Board reported in March 2026 that the federal government’s General Services Administration carries a deferred maintenance backlog of about $50 billion — more than twice its own previous highest estimate — against a portfolio replacement value of around $160 billion. Its annual maintenance appropriation works out to roughly 0.375 percent of that value, against the 2 to 4 percent benchmark. The backlog is what a decade of the gap looks like.

Planned versus reactive

The Department of Energy’s Operations & Maintenance Best Practices guide, prepared by Pacific Northwest National Laboratory, puts the average American facility at more than 55 percent reactive maintenance — run it until it breaks. Its estimate for moving to a proper preventive programme is a 12 to 18 percent cost saving, and it notes many facilities currently purely reactive could save considerably more.

The same guide describes the trap in a sentence worth quoting:

During the time we believe we are saving maintenance and capital cost, we are really spending more dollars than we would have under a different maintenance approach … while waiting for the equipment to break, we are shortening the life of the equipment resulting in more frequent replacement.

US Department of Energy / FEMP, O&M Best Practices Guide, Release 3.0

A real example, on a roof

The clearest case study in that guide concerns a state agency in the northeast with a 360,000 square foot roof, over 22 years old and leaking. Replacement quotes came in between $2.5 and $3 million. Instead they inspected it and repaired what the inspection found — 1,208 square feet the first year at $20,705, then 1,399 square feet the next at $18,217. Total roof cost over six years came to under $60,000.

The roof still had to be replaced eventually. But the six years bought by inspecting and repairing cost about two percent of replacing it early.

Which parts of a building are consumables

The useful way to think about a building envelope is that a few short-lived materials protect several long-lived ones. The Department of Housing and Urban Development publishes estimated useful lives for exactly this kind of planning:

More for property owners