Charted: US Aggregate Tonnage Fell While Real Construction Put-in-Place Kept Rising

Census construction put-in-place (CIP) prints look like a mid-decade boom. USGS quarry ledgers do not. In 2024, US construction aggregates produced for consumption fell about 6% to 2.38 billion metric tons — crushed stone to 1.48 Gt (−5%) and construction sand and gravel to 900 Mt (−7%), per USGS Mineral Industry Surveys [USGS MIS Q4 2024]. Over the same 2019–2024 window, the calendar average of Census C30 total construction spending (FRED TTLCONS) rose roughly 61% in nominal dollars [Census C30]. Strip out construction-cost inflation with a 2019=100 deflator path and real CIP still rises about 15%. Tonnage ends below 2019. That is the story: tonnage vs real CIP, not “America stopped building.”

The interactive dashboard above indexes USGS tonnage against nominal and constructed real CIP, tracks Mt per $1B of real CIP, maps 2024 crushed-stone output by Census division, shows the MCS end-use mix, and scatters unit value against intensity. Toggle Tonnage vs CIP index, Intensity + gap, Division map, Crushed-stone uses, and Price vs intensity; switch the commodity among total aggregates, crushed stone, and sand & gravel; flip real vs nominal CIP on the dual-index panel.

Scoreboard: dollars, tons, and intensity

The intensity ratio is the cleanest single sentence. In 2019 the US moved about 1.76 Mt of aggregates for every billion dollars of real put-in-place. By 2024 that figure is 1.49. Real CIP did not collapse; tons failed to keep the same pace.

Nominal CIP invents a boom tonnage never delivered

If you only watch the C30 dollar print, 2021–2024 look like a construction super-cycle. Nominal CIP’s 2019=100 index climbs into the low 160s by 2024. Aggregate tonnage’s index sits near 97–103 for most of the window and finishes 2024 around 97.5. That wedge is mostly price: materials, labor, and contractor margins inflated the dollar series far faster than physical volume.

Constructed real CIP — nominal CIP divided by a Turner/ENR-style cost deflator anchored at 2019=100 — tells a quieter story. Real put-in-place drifts from 100 in 2019 to about 115 in 2024. That is still growth. It is not a 60% volume boom. Desks that treat nominal CIP as a tonnage proxy will keep over-ordering the recovery.

The 2024 cut was real on the quarry side

USGS MIS annual estimates put crushed stone at 1.48 Gt in 2024 after 1.55 Gt in 2023, and sand & gravel at 900 Mt after 967 Mt. Total construction aggregates: 2.38 Gt after 2.52 Gt [USGS MIS Q4 2024]. MCS salient statistics for crushed stone show a parallel path — sold-or-used near 1.46–1.55 Gt across 2020–2023, with a 2024 estimate around 1.50 Gt — and note weather, high mortgage rates, and softer residential demand as the demand brakes [MCS 2025]. The MIS sample survey is the sharper annual production-for-consumption print used here for 2024; Minerals Yearbook finals can still revise.

Sand & gravel fell harder than crushed stone in percentage terms. That matters because the two products are imperfect substitutes: crushed limestone and granite dominate many eastern and southern pavement markets, while sand & gravel still carry large western and alluvial corridors. A national aggregates total hides that mix shift.

Intensity is the gap that keeps widening

Index the two physical clocks the same way. Aggregates (2019=100) finish 2024 near 97.5. Real CIP finishes near 114.8. The index gap — real CIP minus aggregates — opens from roughly flat in the mid-2010s to about 17 points by 2024. Intensity (Mt per real CIP billion) falls from 1.76 → 1.65 → 1.49 across 2019, 2023, and 2024.

Why would real dollars rise while tons soften? Several honest mechanisms can coexist:

1. Mix shift inside CIP. Manufacturing plants, data-center shells, and high-finish nonresidential work can carry more dollar volume per ton of aggregate than lane-miles of asphalt.
2. Price without volume. Even after a cost deflator, residual price and scope inflation can leave “real” dollars ahead of quarry shipments.
3. Stock and recycle. Stockpiles, recycled asphalt/concrete (still a small national share in MCS notes), and import wedges can buffer local shortages without lifting domestic mine output.
4. Use composition. Only about 72% of crushed stone is construction aggregate in the MCS 2024e use mix; cement and lime claim 17% and 6% [MCS 2025]. Treating every crushed-stone ton as pavement feedstock overstates the road ledger.

None of those excuses a desk that quotes nominal CIP growth as proof that aggregates “kept up.” They did not, on either the MIS tonnage print or the intensity ratio.

Geography: the South still ships; the Midwest cut harder

MIS 2024 crushed-stone production for consumption remains a southern story. South Atlantic alone is about 341 Mt (~23% of the US total). West South Central adds about 268 Mt. Together with East South Central, the South is roughly half of national crushed-stone tonnage. East North Central is still large (~231 Mt) but posted a mid-single-digit percentage drop; West North Central fell harder (~−13% YoY on the division estimate) [USGS MIS Q4 2024]. Mountain divisions also printed double-digit percentage declines from a smaller base.

That geography matters for freight and ready-mix desks. A national −5% crushed-stone print can feel like a deep cut in the Plains and only a soft patch in Florida–Georgia–Carolinas pavement corridors. The dashboard’s division panel and region filter exist so those two experiences are not averaged into one false calm.

Prices rose while the intensity ratio fell

MCS average unit values for crushed stone climbed from about $12.69/t in 2020 to an estimated $17.50/t in 2024; construction sand & gravel moved from about $9.95 to $13.90. The price path and the intensity path move in opposite directions on the scatter panel: higher quarry dollars, fewer tons per real CIP dollar. That is consistent with a market where local zoning, haul distances, and fuel costs push prices even as residential pavement demand softens.

It is not evidence that the US is “out of stone.” USGS repeatedly notes that stone resources are ample nationally while local shortages near metro edges are structural — permitting and land-use conflict, not geology. Price is the local scarcity signal; national tonnage is the volume signal. Conflating them produces bad forecasts.

Caveats (read before you trade the gap)

• Real CIP is constructed. This post deflates FRED TTLCONS annual averages with a 2019=100 construction-cost path. It is not Census’s published constant-dollar VIP table. Treat the level as a companion index for intensity math, not an official BEA or Census real-volume release.
• MIS vs MCS. 2024 annual aggregates here use MIS production-for-consumption estimates (1.48 / 0.90 / 2.38 Gt). MCS sold-or-used estimates can differ slightly; both revise.
• CIP is not aggregate demand. Residential, highway, manufacturing, and power work have different tons-per-dollar. A CIP mix shift can move intensity without any quarry “failure.”
• Not all crushed stone is roads. Cement and lime feedstock sit inside the same MCS ledger.
• Unit values are averages. Urban fringe markets can clear far above the national MCS mean.

What the dual clocks say for 2025 desks

If highway and IIJA-era public work continue to spend while single-family starts stay rate-sensitive, expect dollars and tons to keep disagreeing. Watch three prints together: (1) USGS quarterly MIS aggregates, (2) Census C30 nominal CIP, and (3) an explicit real or volume companion — never nominal alone. The 2019–2024 intensity drop from 1.76 to 1.49 Mt per real CIP billion is the quantitative form of that disagreement. Closing the gap would require either softer real put-in-place or a clear rebound in quarry shipments. 2024 delivered neither on the aggregates side.

For desks that still need one line: US aggregate tonnage did not keep up with real construction put-in-place — and the Mt-per-real-dollar gap is wider than it was before the pandemic.