A Government Accountability Dashboard. Measure outcomes, not promises.
The goal is not to judge intent or ideology. It is to answer one question: did the indicators improve after the money was spent and the policy was implemented?
Start with Canada, or drill into a province where provincial data exists. When a metric has no provincial series, the national figure is shown and tagged.
Tile-grid layout (a cartogram). A geographic MapLibre choropleth is the planned upgrade. Greyed tiles have no data in this prototype.
Now viewing: Canada (national)
Provincial coverage is strongest for health spending, ER waits, family-doctor access, surgical waits, crime severity, and rent/vacancy (by major-metro proxy). Homelessness, home prices, and policing dollars are national-only here.
For each sector: what was spent, what was measured, and a plain-language verdict on whether outcomes improved. Open "Evidence trail and caveats" under any card to see every source.
One formula lets housing, health, and safety be compared on the same footing. It rewards measured improvement, not activity.
When spending rises and outcomes worsen, efficiency is negative: more money, less result. The bars below show, for Canada, how the spending change compares with the outcome change in each sector.
Outcome change is the average signed movement of that sector's outcome indicators, in the direction that counts as improvement (for example, shorter ER waits and lower crime severity are improvements). It is a summary, not a substitute for the individual rows above. Housing spending growth is shown qualitatively because federal housing dollars are reported as overlapping multi-year envelopes that do not reduce cleanly to one growth rate.
Accountability lives in the gap between a policy being announced and an outcome actually moving. A real version of this dashboard plots each step on a timeline so you can see whether the indicator changed before or after the intervention. Illustrative example, federal housing:
Not the number of programs, press releases, or meetings. The measured change in the thing the program was meant to fix.
Every figure links to the original dataset, the source body, the year, and a confidence rating. No black box.
Gaps, proxies, methodology breaks, and forecast-not-actual figures are labelled, not hidden. Correlation is never dressed up as cause.
The same efficiency framework spans housing, health, safety, and beyond, so sectors can be compared fairly.