Construction Cost Indices

Construction cost indices, the published series an estimator reaches for before anything about a project is drawn, have a geography of their own, and Michael Stern’s subject, the pricing of buildings that have few precedents, begins where that geography runs out. Every index is anchored to particular places and a particular basket of goods, and each is honest about its limits in a way that its users often are not. This page walks through the main ones, from the twenty cities where one magazine’s reporters price steel each month to the single place no index visits, which is the building unlike the ones in its sample.

Twenty cities and a basket

The oldest stop is Engineering News Record. The magazine’s guide to its own indexes describes price reporters in 20 American cities who quote the same suppliers each month for a fixed basket: 25 hundredweight of fabricated structural steel at the 20 city average, 1.128 tons of bulk portland cement and 1,088 board feet of 2 by 4 lumber priced locally, plus a labor component.¹ That labor component is the only difference between its two series. The Construction Cost Index carries 200 hours of common labor; the Building Cost Index carries 68.38 hours of skilled labor averaged across bricklayers, carpenters and structural ironworkers, and is the more applicable of the two to structures.¹ Both measure what the hypothetical package costs against a base set so that it cost $100 in 1913.¹

Figure

ENR's two indexes compared

	Construction Cost Index	Building Cost Index
Labor component	200 hours of common labor at the 20 city average wage and fringe rate	68.38 hours of skilled labor, averaged across bricklayers, carpenters and structural ironworkers
Materials component	25 hundredweight of fabricated structural steel, 1.128 tons of bulk cement, 1,088 board feet of 2 by 4 lumber	The same basket
Best use	Work where labor is a high share of cost	Structures
What it is not	A measure of the gap between cities	A seasonally adjusted series

The magazine’s cautions are as useful as its numbers. The indexes measure a trend in one city or in the nation as a whole and not the cost gap between cities, which the guide calls one of the more common errors in their use; they are not seasonally adjusted, so wage increases that land once or twice a year, cement prices that move in spring and lumber prices that swing month to month can mislead anyone reading less than twelve months of data; and the 20 city average, with more elements and a smoother trend, is generally preferable to a single city’s series.¹

Figure

Reading an index without being misled

1. 01
   Not a differentialThe indexes track the trend in one city or the nation; they do not measure the gap in cost between two cities.
2. 02
   Not seasonalWages move once or twice a year, cement in spring, steel monthly, lumber most of all; a reading over less than twelve months can miss the pattern.
3. 03
   Prefer the averageThe 20 city average has more elements and a smoother trend; a single city is prone to spikes.
4. 04
   Watch the wage calendarA stalled labor negotiation can hold an old rate in place and make inflation look lower than it is.

A contractor’s own job sites

The second stop is inside one builder’s books. Construction Dive reports that the Turner Building Cost Index, a series with an 80 year history, is compiled from the firm’s own projects nationally, folding in labor rates, productivity, material prices and how competitive local conditions are, which is why it may not match other published indices.² In the third quarter of 2024 it rose 3.64 percent year over year while federal statistics showed nonresidential input prices slightly below a year earlier, a gap the index’s compiler put down to localized activity and mega projects in particular regions.²

The readings move with the market. An owners’ trade publication records the index at 1552 in the second quarter of 2026, up 1.44 percent on the quarter and 5.15 percent on the year, with data centers and semiconductor plants sustaining demand and shortages of skilled mechanical and electrical labor named as the main constraint.³ A decade earlier an infrastructure journal noted the same index at 1006 for the last quarter of 2016, up 4.90 percent on the year, with the same complaint about skilled labor and the compiler’s standing caveat that the index is nationwide and may or may not reflect regional conditions in any given quarter.⁴

Sixteen cities, every quarter

The third stop is a quantity surveyor’s survey. Rider Levett Bucknall’s North American report for the first quarter of 2025 charts construction cost trends in 16 cities, newly including Miami and Nashville, and found a national average quarterly change of 0.98 percent, down from 1.11 the quarter before, with the year over year rate at 4.35 percent against 5.86 a year earlier.⁵ Boston, Chicago, Honolulu, Portland and Seattle rose faster than the national figure that quarter; Denver, Las Vegas, Los Angeles, Nashville, Miami, Phoenix, San Francisco and Washington rose more slowly.⁵ The firm’s president read the data as showing labor cost, rather than tariff headlines, as the real pressure.⁵

Eight regions and one model building

The fourth stop prices a building that does not exist. Mortenson describes its index as calculated quarterly by pricing a representative nonresidential project in cities across eight regions.⁶ For the first quarter of 2026 the national series rose 1.58 percent on the quarter and 5.6 percent over twelve months, but the city readings scattered: 7.87 percent in Salt Lake City over the year, 6.84 in Denver, 5.39 in Chicago and 3.19 in Portland, with previously negotiated union agreements, mega project demand for electrical capacity and metal intensive scopes, and tariffs and energy costs named as drivers.⁶ The same model building, priced in different places, produces the spread that the national number hides.

Figure

Twelve month change in one contractor's index, by city

• Salt Lake City7.87
• Denver6.84
• Minneapolis6.66
• Phoenix6.32
• Milwaukee6.22
• Chicago5.39
• Seattle3.48
• Portland3.19

percent

The composite buildings of the older indexes

The fifth stop is the family of indexes built around a fictional structure. A civil engineering reference surveys them. Boeckh’s index gathers data from 205 American cities in three categories, materials, labor, and tax and insurance, with both union and open shop wages considered. The Lee Saylor index weights labor at 54 percent and materials at 46, averaging 23 materials across 20 cities and nine trades in 16, on a 1967 base. The Means historical index prices a hypothetical composite building made of nine structure types, six equipment rentals, 21 trades and 66 material components, each weighted by how much the fictional structure needs.⁷ Every one of these is a different imaginary building, and the differences between them are a first measure of how much the choice of building matters.

The producer’s gate

The sixth stop is upstream of the site. A construction economist’s annual inflation note separates series that are routinely confused. The consumer price index tracks a basket of household goods and is, in his words, not related at all to construction. Producer price indexes for construction inputs track material cost at the producer’s gate, not prices or bids at the as built level. General indexes such as ENR’s Building Cost Index and the RSMeans index are widely used to adjust project costs yet do not represent whole building cost either, because neither includes contractor margins.⁸ Only a selling price index, tracking the final cost of a completed building with the margins of the trades and the main contractor inside it, captures the full movement.⁸

Figure

Four series that get confused

1. 01
   Consumer pricesA basket of household goods, with no relation to construction at all.
2. 02
   Producer prices for inputsMaterial cost at the producer's gate, not what is bid or built.
3. 03
   General construction indexesENR's and RSMeans' series, widely used to adjust project cost yet carrying no contractor margin.
4. 04
   Selling priceThe final cost of a completed building, including the margins of the trades and the builder.

The midpoint of construction

The seventh stop is a date rather than a place. The same economist explains that a budget prepared from current costs for a project whose spending midpoint is two years away must carry an inflation factor to that midpoint, because half the cost falls before it and half after, and that any project put on hold must have its inflation recalculated from the new dates.⁸ He adds the observation that explains most of the regional scatter above: the level of activity in an area is one of the best predictors of construction inflation, since contractors competing for scarce work lower their bids and contractors with more work than they can take raise them.⁸ Since 2011 his nonresidential building inflation has averaged 4.8 percent a year including the unusual years of 2021 and 2022, and 3.8 percent without them.⁸

The clause that reads the index

The eighth stop is a contract. A standard forms publisher describes a material price escalation clause that moves the contract sum, up or down, once a named index crosses a threshold, and calls choosing a reliable index that actually tracks the materials in a given project one of the harder parts of drafting it.⁹ On a lump sum contract the builder and its suppliers otherwise carry price increases that are not their fault, which the publisher lists alongside shorter bid validity periods, cost reimbursable agreements, phasing, early procurement and larger contingency as ways to share the exposure.⁹ How such a clause narrows the spread of prices is taken up under bid dispersion.

The building no index has priced

The last stop is the one that matters for this site. Every series above prices a standard thing: a fixed basket, a representative project, a composite structure, a model building. None contains a tower of unusual slenderness, a facade of untried assembly or a core climbing at a pace the local trades have not worked before, because an index is an average and the unusual building is, by definition, off its sample. What an index supplies to such a building is the movement of the inputs it shares with everything else, the wage, the cement, the steel, escalated to mid construction as the economist advises.⁸ What it cannot supply is the premium the building carries for being itself, which the cost estimating page builds from components, the skyscraper economics entry traces riser by riser, and the underwriting pillar treats as the whole problem.