Cement Production, Shipments and Stocks (Table 16100042)

Every construction cycle announces itself first in tonnes. This table is Statistics Canada's monthly record of the Canadian cement economy's physical flow: 2,067 data series across 159 consecutive monthly periods from January 2004 through March 2017, every reading stated in plain metric tonnes.

Four activities make up the ledger - production, shipments, stocks at the beginning of the month, stocks at the end of the month. Four product lines slice each activity: total cement products, Portland cement, masonry and other cement, and clinker. Sixteen member combinations per month, and the construction arithmetic ties them together: opening stock plus production minus shipments should land near closing stock, so internal consistency can be tested rather than assumed.

The table carries a stable identity - Table 16100042 - that every figure traces back to, which matters when a claim needs a citation sturdier than the current news cycle. Get a sample of this dataset and the first rows arrive with your product lines and year range already applied.

One row per activity-product-month combination, flat and immediately pivotable. Illustrative rows in the delivered column order:

REF_DATE : 2004-01    GEO     : Canada
activity : Production           product: Total, cement products
unit     : Metric tonnes        vector : v29505576
VALUE    : 618671

REF_DATE : 2004-01    GEO     : Canada
activity : Production           product: Portland cement
unit     : Metric tonnes        vector : v29505585
VALUE    : 545560

REF_DATE : 2004-01    GEO     : Canada
activity : Shipments            product: Portland cement
unit     : Metric tonnes        vector : v29505586
VALUE    : 433942

Those are the January 2004 openings of the series: 618,671 tonnes of total cement products produced, 545,560 tonnes of that Portland, and 433,942 tonnes of Portland shipped out the gate in the same month. The VECTOR column is worth noticing - each line of the sixteen-member grid carries its own stable identifier, so any single series can be tracked across the whole 159-month run without re-processing anything around it.

Twelve documented fields define every row, and eleven of them are typed dimensions or bookkeeping - the twelfth, VALUE, is the tonnage itself. Because each observation is one activity-product-month member, the table pivots cleanly into wide matrices by product or by activity. Sparse flags (STATUS, SCALAR_FACTOR) usually arrive empty, and empty means empty rather than zero - honest gaps you can filter instead of silently averaging away.

Geography - Canada as a single national total. No provincial, territorial or city breakdown lives inside the table; if your question needs sub-national resolution, this record answers the national layer and leaves the regional split to complementary sources.

Temporal - January 2004 through March 2017, monthly, 159 consecutive periods with no gaps. The panel is closed: it captures the pre-crisis boom, the 2008-2009 contraction, and the recovery years in one fixed vintage, so figures never shift under a running analysis.

Granularity - monthly observations crossed by four activities and four product lines, sixteen member combinations per period, all in one unit. That is enough structure to compute inventory turns, shipment-to-production ratios and product-mix shares without any reshaping of your own.

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

Your cadence is your call even though the panel itself is closed: load it once as a backfill, then keep it in rotation beside your live feeds on whatever schedule your models expect. Samples ship in exactly the twelve-field shape shown above.

• Construction demand forecasting - shipments are physical demand, not invoiced revenue: they carry no currency effects and move ahead of price indices, which makes them a clean leading input for Canadian construction outlooks; the wider context sits on our construction materials data hub.
• Inventory-cycle analysis - month-opening and month-closing stock columns turn inventory change into a subtraction instead of a guess, the classic tell of whether producers are accumulating tonnage or draining it.
• Capacity-utilization signals - the spread between production and shipments shows when kilns run ahead of the market; the supply-side view of the same question lives in our global database of cement production assets and upstream suppliers.
• Embodied-carbon baselining - the clinker series isolates the high-emission intermediate before grinding, ready to pair with emissions intensities from our global CO2 emissions from cement production data.
• Backtesting and feature engineering - 159 monthly periods times sixteen members of clean single-unit series drop straight into panels; related modeling patterns sit on our data scientists use cases page.

Data scientists and ML engineers get 2,067 tidy monthly series in one unit - no harmonization chore before the first model fit; see data scientists use cases. Investors and quants get a physical-volume read on Canadian construction that moves independently of reported revenue, useful as a sanity check on company guidance; see investors and quants use cases. Market researchers and consultants get Portland-versus-masonry mix shifts as direct evidence of building-type trends; see market researchers use cases. Competitive-intel teams get inventory behavior by product line as a read on producer discipline. Journalists, academics and students get a citable national series with a stable table identity behind every figure.

Within cement data, this record owns the national physical-flow ledger for Canada: four activities, four product lines, one country, one unit. Global production aggregators such as our monthly global cement production data carry Canada as one line among dozens of economies, and asset-level databases know which plants exist but not what those plants shipped each month. Layered together they answer three different questions - who makes cement, how much the world made, and what Canada's own tonnage did month by month. The full slate sits on our best construction materials datasets ranking.

Three things. First, the panel closes at March 2017: treat it as a fixed historical window, and pair it with a still-growing series when your question extends past that date. Second, geography is national only - Canada as a single economy, with no provincial or city detail inside the table. Third, every value arrives in metric tonnes with a status flag beside it, so flagged observations can be filtered out rather than silently folded into averages.