A standing indicator

How much of what Canada asks about itself can be answered

Across every source in this warehouse, 8,639,520 questions could have had an answer. 82.2% of them did.

82.2% answered 9.5% too small to report 8.3% withheld for another stated reason

Where the effect falls

The same rules, very different results

ON gets an answer to 91.7% of what is asked of it. NU gets an answer to 65.9%, a gap of 26 percentage points. No rule was applied differently in the two places. The population is smaller, so more cells fall below the threshold at which a number can be published safely, and a smaller share of the questions asked about the place can be answered at all.

Every province, territory and the national series, by how much of what was asked came back with a number. Sorted by the share answered, least first.
Place	Questions asked	Answered	Too small to report	Withheld, other reason	Sources
NU	130,307	65.9%	24.8%	9.2%	117
NT	139,332	71.0%	21.8%	7.2%	116
YT	202,869	71.4%	16.5%	12.0%	124
NL	1,687,195	75.0%	2.8%	22.2%	138
PE	349,634	80.0%	15.5%	4.5%	138
NS	973,117	80.3%	5.0%	14.7%	224
MB	419,864	82.2%	9.6%	8.2%	135
BC	2,036,933	84.1%	15.2%	0.6%	147
NB	366,506	84.4%	12.0%	3.5%	137
SK	386,485	84.7%	11.5%	3.8%	132
AB	399,636	85.7%	11.1%	3.2%	139
QC	454,974	90.1%	6.7%	3.2%	135
ON	503,618	91.7%	5.5%	2.8%	138
CA	392,916	94.7%	2.6%	2.7%	139

whomatters.ca answerability-by-place whomatters warehouse, dataset cells=8687740;computed=2026-08-14, computed 2026-08-14 Roll-up rows only; a breakdown row describes the same cells at a finer resolution and is not added in.

97 further place codes appear in the warehouse below the province line, published by single sources at regional or sub-provincial resolution. They are held out of this table because a reader would take them for peers of a province, and they are not. The full place list is here.

Over time

The series

Every measurement of the indicator. One row per distinct warehouse; a rebuild that changes nothing does not add a row.
Measured	Warehouse	Cells	Answered	Too small	Movement	Verdict
2026-08-13	`cells=7146622;computed=2026-08-12`	7,146,622	80.57%	9.94%	·	FIRST
2026-08-13	`cells=8687740;computed=2026-08-13`	8,687,740	82.18%	9.54%	-0.40 pts	WAREHOUSE GREW
2026-08-14	`cells=8687740;computed=2026-08-14`	8,687,740	82.18%	9.54%	0.00 pts	UNCHANGED

How this was measured

A question is a cell that could have had an answer: one measure, one place, one period, one breakdown. Cells the publisher marks not-applicable, and place-and-measure combinations the publisher does not offer a row for at all, are excluded from the denominator; counting them would inflate every rate and make a well-designed table look evasive. That exclusion is 48,220 cells out of 8,687,740.

Totals are taken from the roll-up row for each source and place. The same table also holds a row per breakdown value inside each source, describing the same cells at a finer resolution; adding those to the totals would count every cell about 2.2 times. The exporter's selftest proves that the wrong basis is wrong rather than leaving it to a comment.

Computed by warehouse/build-metrology.mjs, which reads canonical read-only and reconciles its output against it. Published by warehouse/export-metrology.mjs.

The whole measurement layer is published as data. metrology.json carries every figure on this page plus the per-breakdown detail behind it: answerability for each source, place, dimension and dimension value, and the instrument vector for every subject. It was generated and deployed before anything linked to it, which is the ordinary way a measurement goes unread. answerability.json is the smaller per-source cut the explorer fetches to annotate a chart.