American Cement Association Market Intelligence and Forecasts

American Cement Association Market Intelligence and Forecasts is the demand-side reference set for US cement, produced by the economics team of the American Cement Association - the Skokie-based trade association that spent a century as the Portland Cement Association before rebranding. Where government statistics count what was shipped or spent, this program asks the narrower question the industry actually prices against: how much cement will be consumed, where, by which part of construction, over what horizon.

The answer arrives through a family of products rather than one table. The U.S. National Forecast projects consumption five years out, combining macroeconomic drivers of construction with proprietary cement consumption per construction dollar modelling, re-issued each spring, summer and fall. The State Forecast extends the same horizon downward, projecting constant-dollar put-in-place (PIP) spending across 16 construction sectors plus more than 30 additional variables per state. The Quarterly Survey of Portland Cement Consumption measures volumes across 18 user segments with regional distributions and projections. Apparent Use of Portland Cement by State & Market estimates consumption across 46 construction markets for 56 states and geographic areas. The State and Market Trend Report holds twenty years of history per state, and the U.S. Cement Industry Annual Yearbook packs over fifty pages of tables - economic activity, construction drivers, consumption, infrastructure, supply, operating characteristics, international markets.

Datadory delivers the entire program as one normalized feed: same field spine whether a row originates in a five-year forecast or a quarterly survey. Where it sits among its peers is mapped on our broadline retail data hub.

Four shapes cover the family - forecast rows, apparent-use rows, survey rows and the national headline:

# Apparent Use of Portland Cement by State & Market
product       : Apparent Use by State & Market
market        : Highway and paving        # 1 of 46 construction markets
geography     : Texas                     # 1 of 56 states and areas
measure       : apparent cement consumption, tonnes
period        : FY2025

# State Forecast - one row per state x sector x forecast year
product       : State Forecast (spring edition)
geography     : California
sector        : Office                    # constant-dollar PIP spending, 1 of 16
forecast_year : 2027

# Quarterly Survey of Portland Cement Consumption
product       : Quarterly Survey of Consumption
user_segment  : Ready-mix concrete        # 1 of 18 segments
region        : Midwest
quarter       : Q2 2026

# U.S. National Forecast - headline row
product       : U.S. National Forecast (spring)
measure       : total US portland cement consumption
forecast_year : 2028
model_basis   : cement consumption per construction dollar

Read the coordinates, not the blanks: every row is pinned by product, geography, market-or-segment and period, which is what lets one delivery pipeline carry forecast rows and survey observations side by side without reshaping either. Live values ship with the sample, cut to the markets and states you name.

Six documented fields form the working dictionary, and they divide cleanly into coordinates and cargo. product_series names which of the six-plus products a row came from - the join key that keeps forecast rows from masquerading as survey observations. geography resolves against the 56-member state-and-area list or a named survey region. market_segment holds one of the 46 construction classifications used in apparent-use and trend-report series, or one of the 18 user segments from the quarterly survey. period_or_forecast_year distinguishes a survey quarter from a year inside the five-year horizon - the two time axes run in parallel across the family. The cargo fields are cement_consumption_volume (tonnes, estimated or projected) and pip_spending (constant-dollar put-in-place value, 16-sector split).

Everything beyond that spine folds under additional fields on request, because the association publishes product scopes rather than data dictionaries: the yearbook's 50+ table inventory, monthly Monitors and Flash Reports from the Monitor Outlook Plus bundle, competitive-materials comparisons, per-segment regional distributions, and the vintage stamps separating spring from summer from fall editions. Name the product lines when you request a sample and the relevant extensions arrive documented against delivered extracts.

Five families sit outside the core six, mapped to your scope at sampling:

• Yearbook table inventory - the economic-activity, construction-driver, infrastructure, supply, operating-characteristics and international-market tables behind the 50+ page annual volume.
• Monitor Outlook Plus bundle content - monthly US Monitors with data tables, Flash Reports, monthly Tracking Reports and monthly Competitive Materials reports.
• Segment-level regional distributions - the within-segment geographic splits the quarterly survey publishes alongside its projections.
• Full trend-report matrices - twenty years of consumption across all 46 classifications per state, plus PIP spending for 23 categories.
• Forecast vintage metadata - spring, summer and fall editions identified separately, so revision analysis can compare successive vintages of the same outlook.

The pattern is deliberate: documented-on-delivery beats guessed-at-up-front. The six-field core describes what every record carries; these extensions describe what specific product lines carry once extracted.

Geography is the United States end to end. The National Forecast speaks nationally; the State Forecast works state by state; the apparent-use series spans 56 states and geographic areas - a list broader than the 50 states, capturing territories and district-level markets; the quarterly survey distributes each segment across named regions.

Temporal depth runs in both directions. Forward, the National and State forecasts project a five-year horizon, re-estimated three times annually as construction conditions move. Backward, the State and Market Trend Report holds twenty years of history per state, and the yearbook refreshes annually. Between the two sit the monthlies - Monitors, Flash Reports, Tracking Reports - that track the current year as it unfolds.

Granularity is the program's real differentiator: 46 construction markets crossed with 56 geographies yields thousands of distinct demand cells in the apparent-use series alone, the 16-sector PIP split decomposes each state's construction spending, and 18 user segments resolve the quarterly survey. That resolution is why teams pair it with coarser national series rather than replacing them - see our building products data guide for how the pieces fit.

API, files, or your warehouse. Daily, weekly, or hourly.

Your cadence is your call even though the underlying publications move on their own calendar - forecasts land three times a year, the survey quarterly, the yearbook annually. Backfill the twenty-year trend-report history once, then let each new vintage rotate into place on whatever schedule your models expect. Samples ship in exactly the six-field shape documented above, cut to the markets, states and product lines you care about.

• Demand forecasting for cement and concrete - build multi-year volume models on the national outlook, then calibrate state exposure with the 16-sector PIP projections; producers run this loop ahead of every capacity decision.
• Capacity and network planning - terminal, grinding and distribution decisions keyed to per-state consumption trajectories rather than national averages that hide the growth pockets.
• Market sizing for building products - apparent-use tonnage across 46 markets gives adjacent-materials players a defensible denominator for share estimates; see market-sizing patterns.
• Substitution monitoring - competitive-materials reporting tracks where cement loses ground to asphalt, steel and timber, segment by segment.
• Infrastructure investment signals - highway-and-paving classifications flag where publicly funded concrete demand lands next, state by state.
• Equity and credit research - consumption trends plus yearbook operating-characteristics tables give analysts demand and supply sides in one package; see quant-backtesting patterns for the modeling angle.

Data scientists and quant researchers get the rarest thing in construction data: a forward-looking panel with stated model basis and fixed release rhythm, delivered as coordinates instead of scanned exhibits - joinable against housing starts and Census construction-spending series without manual transcription.

Strategy and corporate development teams at cement producers, aggregates groups and building-products manufacturers get per-state demand projections to pressure-test expansion theses before committing capital.

Market researchers and consultants get citation-grade apparent-use numbers across 46 markets, with twenty years of state history already assembled - the difference between quoting a number and defending one.

Investors and analysts treat cement consumption as the classic read on construction activity itself: a coincident indicator for materials equities and a leading one for downstream building trades.

Policy and infrastructure analysts connect public capital programs to physical material requirements through the same state-level cells everyone else uses.

Within our catalog, three neighbors bracket this record. The AGC Construction Data Hub tracks roughly 100 producer-price series for construction inputs including cement - price, not consumption, and no forward view. The Statistics Canada building construction price indexes own whole-building cost measurement, but for Canada and again without a demand forecast. The APA engineered wood collection mirrors this program's structure for a different commodity - association-run, mill-reported, forecast-bearing - but for structural panels rather than binders.

This dataset is the only one of the four that projects who will consume how much of one material, where, five years out. Its limits mirror its strengths: one country, one material, granularity bounded by the association's own 46-market and 56-geography frames. Pair it with price-series feeds for the cost side and cross-border trackers for the global picture, and each covers what the others concede.

Three things. First, scope it: name the product lines - national forecast, state forecast, quarterly survey, apparent use, trend report, yearbook - and the markets and states you track, because the family is broad enough that an unscoped sample wastes everyone's time.

Second, mind the inference boundary. Product descriptions document scope precisely - 46 markets, 56 geographies, 18 segments, 16 sectors - but not column layouts. Our six-field core reflects published structures; anything finer is documented against delivered extracts, so validate parsers against the sample rather than against assumptions.

Third, respect the cadence asymmetry: forecasts re-issue three times a year and the survey quarterly, but the yearbook and trend reports move annually or slower. Plan pipelines around the fastest series you actually consume rather than the whole bundle. Samples precede any commitment, and the schema in the sample is the schema you integrate against.