Methodology

How every published number is computed, and where it comes from. The methods below are the same for every region MetroTrack covers; the specific agencies and sources are listed per region at the end.

How each metric is computed

Job-access score

The signature metric: how many jobs you can reach by transit, from where you are, in a reasonable time.

access_score(origin, cutoff) = Σ jobs(d) for every destination block d reachable from origin by transit within cutoff minutes

Origins are H3 hexes for the ambient map layer, or any clicked point on demand.
Cutoffs are 15 / 30 / 45 minutes, departing at a user-selectable time (default weekday AM peak).
Walk access/egress is capped at ½ mile to and from a stop.
Jobs come from Census LEHD LODES (block-level workplace counts); transit legs from GTFS.

Full modeling notes ↗

Service-delivery shortfall

The hiring pillar's headline: staffing shortage shows up as service the authority scheduled but didn't deliver.

shortfall% = (scheduled_trips − delivered_trips) / scheduled_trips, per authority over a period

Scheduled service comes from GTFS static; delivered service from GTFS-realtime and published delivered-service reports.
Reported as a percentage (comparable across authorities) plus absolute trips or hours lost.
Authority-wide headline, with route-level and mode (bus / rail) drill-down where data allows.
Corroborated by a secondary posting-based vacancy series (open reqs) scraped weekly.

Full modeling notes ↗

Funding: budget vs. actual

Where the money goes — budgeted operating expense against audited actuals, per service board over time.

For each authority × year: budgeted operating expense vs. actual (audited) operating expense

Budgeted figures come from the RTA adopted operating budget (budget book / quarterly reports, transcribed with citation).
Actuals come from the FTA National Transit Database (audited operating expense, fare revenue, unlinked trips by agency × year).
Historical financials and ridership are cross-checked against RTAMS.

Full modeling notes ↗

Usage mapping

Routes and stops mapped against the people and destinations they serve.

Population and jobs are binned to H3 hexes; routes/stops overlaid with a walkshed to nearby destinations

Jobs from Census LEHD LODES (block-level); population from the 2020 Census centers of population.
Destinations (shops, airports, attractions, venues) from Overpass / OpenStreetMap.
Routes, stops, and schedules from GTFS static for CTA, Pace, and Metra.

Full modeling notes ↗

Regional data & parameters — Chicagoland

The methods above apply to CTA, Pace, Metra, and the NITA (formerly RTA). Jobs are sourced from Census LEHD LODES8 (workplace area, Cook County and the collar counties). Funding figures come from the RTA adopted operating budget and the FTA National Transit Database (CTA / Metra / Pace subset).

Career-page snapshots are collected weekly from each authority's own ATS (CTA via Taleo, Metra via Cadient, Pace via its careers site).
The ambient job-access layer currently uses a straight-line walkshed; openrouteservice network isochrones activate once an API key is configured.
The NITA funding-reallocation dollar amount is a placeholder until verified against primary legislation.

Dig deeper

Data-source ledger ↗Source on GitHub ↗About the project →