How much does foundation repair cost?

• There is no NAICS code for foundation repair, and we checked properly this time: the 2022 NAICS index has 20,398 entries and the phrase appears in none of them. The government files the work by METHOD, and the methods live in six different industries. Verbatim from the index: helical piers under SITE PREPARATION contractors (238910), underpinning and shoring under ALL OTHER SPECIALTY TRADE (238990), urethane slabjacking and epoxy injection under OTHER FOUNDATION, STRUCTURE AND BUILDING EXTERIOR (238190), foundation dampproofing and waterproofing under BUILDING FINISHING (238390) next to the caulkers, rebuilt footings and concrete repair under POURED CONCRETE (238110), and block or stone walls under MASONRY (238140).
• The neatest proof that the classification follows the MATERIAL rather than your problem: a sunken slab. Lift it by pumping urethane foam underneath and the NAICS index calls you a 'Urethane slabjacking contractor' (238190, billing $111.58 a field hour). Lift the same slab the same inch by pumping a cement slurry underneath and the index calls you a 'Mud-jacking contractor' and files you under poured concrete (238110, billing $100.03). Same job, same outcome, $11.55 an hour apart, because of what is in the hose.
• Nationally the six trades bill within 1.20x of each other: $97.27 a field hour for masonry, $100.03 for poured concrete, $111.58 for slabjacking, $115.11 for waterproofing, $115.72 for underpinning, $117.15 for piers. All measured from the 2022 Economic Census, in 2025 dollars.
• But do NOT read that 1.20x as a limit on what two contractors can charge you, and we originally did, which was a fallacy of division. Those are six national averages. Go to state level in the same Census file and a SINGLE trade's billed rate spans 2.0x to 2.8x across the country. Pooled across all six trades and all states, the observed range is $60.38 to $188.88 a field hour, a 3.13x spread, and those are still averages of firms rather than firms.
• What that does give you is a CEILING, and it is the useful thing here. $188.88 a field hour is the highest billed rate the Economic Census observes for any of these six trades in any state in the United States. Scope, by contrast, is bounded by nothing and measured by nobody. So if your quote implies $379 a field hour, the excess over the ceiling is not the price of labour, and nobody has told you what it is.
• American Housing Survey 2023: 5.3% of American HOUSES have holes, open cracks or crumbling in the foundation. That is 5.6 million of them. Note the denominator: the survey skips the whole exterior-condition module for every unit in a building of two or more apartments, so the answering universe is 106.2 million of the 145.3 million US housing units. Measured against all housing units the figure is 3.9%. We first published this as '5.3% of US homes', which was a unit-of-analysis error.
• The age of the house predicts foundation trouble far better than the foundation does. 10.0% of pre-1950 houses have foundation cracks or crumbling, against 6.4% of 1950s and 60s houses, 4.7% of 1970s and 80s, 3.3% of 1990-2009, and 2.4% of houses built since 2010. A four-fold gradient.
• The crawl-space scare is mostly an age artefact, and it is the trap this page exists to spring. Raw, crawl spaces look far worse than slabs: 6.8% against 4.4%. But crawl-space houses are older, with a median build year of 1970 against 1980 for slabs and 16.6 years between them on the mean. Standardise both onto a common age mix and the gap falls from 2.4 points to 1.0, which is barely significant (z = +2.00, p = 0.046). About 60% of the raw gap is simply that crawl-space houses are older, although the survey's own error bars put that share anywhere between 33% and 87%, so we will not pretend to more precision than that.
• The firm billing $117.15 a field hour for pier work pays the person doing it $35.07. A 3.34x markup, measured, and it is the ordinary economics of a trade with equipment, insurance for undermining somebody's house, and a truck: not evidence that anyone is being gouged.
• Nothing on this page, and nothing on any page, can tell you how many piers your house needs. That is a structural engineering question. Hire an engineer who is NOT also selling you the repair, get the scope in writing, and take that scope to the contractors. Here is the arithmetic behind the advice: the rate is bounded and measurable, and the scope is neither. Whoever decides the scope decides the bill, and right now that person is the man who profits from its size.

What is sourced here, and what is ours. The billed rates are sourced and they are the spine of the page: the Economic Census records what each of the six trades takes in for construction work and the field hours behind it, so we can compute what an hour of crew time on your property bills at, net of materials and subcontracting. The cost mix comes from the same place, and so does the wage, which is how we can tell you that the firm billing $117.15 an hour is paying the man in the trench $35.07. An honest note on those two boxes, because they look like more evidence than they are. The rate and the labour share are not two independent facts. The rate is (receipts minus materials minus subs) divided by hours; the share is that same numerator divided by receipts. Divide one by the other and you get receipts per field hour. So the model rests on ONE measured quantity, and we split it into two boxes only so you can see and override each half. That is presentational. It is not corroboration, and we would rather say so than let the page look twice as sourced as it is. The prevalence is sourced, and it is the other half of the page. The American Housing Survey asks every household in a HOUSE whether there are holes, open cracks or crumbling in the foundation. 5.3% say yes: 5.6 million. We compute the cross-tabs from the microdata ourselves, with the survey's replicate weights, so the error bars are real ones. Ours is the scope, and it is the whole ballgame. The crew size and the days on site are the two numbers that decide your bill, and we cannot know either. We default to three people for two days because that is an ordinary small pier job, but a full perimeter underpinning is weeks. Nothing on this page can tell you how many piers your house needs. Only a structural engineer can, and since the rate is bounded while the scope is not, that written scope is worth more to you than any amount of haggling over the hourly rate. We will not print a price for a repair we cannot source, and here is everywhere we looked. The American Housing Survey asks the cost of 37 named kinds of improvement job (37 defined in the codebook, 36 of which actually occur in the 2023 data: landslide damage never appears). We read all 37, and foundation is not among them, so a foundation repair lands in "other major improvements or repairs inside home" with no sub-code and an uninterpretable median. DOE and NREL's measures database, both workbooks, all 93 sheets: its basement, crawl space and slab sheets are insulation assemblies, and there is no pier, underpinning, slabjacking or crack-injection line item in any of them. The Economic Census gives us receipts and hours but counts no piers, so it cannot price one. Its Kind of Business file, which rescued our gutter page, has 96 categories and no foundation-repair line. And the federal price indexes are indexes: they say foundation work got dearer, never what it costs. If a free source publishes an installed price for a push pier, we have not found it and we would like to. The biggest single assumption, named. The cost mix we use to gross your labour up into a whole job is a TRADE-WIDE average, and those trades do plenty of work that is not residential foundation repair: site preparation contractors dig commercial sites, poured concrete contractors mostly pour new buildings. If a residential repair carries proportionally more steel and less digging than the trade average, our materials line is too low. We flag it rather than bury it, because it is the assumption most likely to be wrong. And a sentence we shipped on this page that was simply false. We wrote, as the loudest line here, that a five-fold gap between two quotes "is not a rate difference and it cannot be", on the grounds that the six trades' national rates span only 1.20x. That is a fallacy of division: a spread of six group averages tells you nothing about the spread between two firms. It is the same error we made seven times on our gutter page. So we went back to the same Census file at state level and measured the dispersion instead of assuming it: within one trade, 2.0x to 2.8x across the states; $60.38 to $188.88 pooled. Rate variation is real. The advice survives, because scope is still unbounded and still unmeasured, but the arithmetic we offered as proof did not, and you should know we got it wrong.

Sources: US Census Bureau, American Housing Survey 2023 National PUF (foundation condition, foundation type and year built, with 160 replicate weights) · US Census Bureau, AHS 2023 Items Booklet (the foundation question wording, and the list of 37 improvement job types that does not include foundations) · US Census Bureau, 2022 Economic Census, EC2223BASIC (the billed rates, wages and cost mixes of the six trades, nationally and by state) · US Census Bureau, 2022 NAICS index file (which shows helical piers, underpinning, slabjacking and mudjacking landing in four different industries) · BEA, residential improvements price index (deflator)