Two indices designed to track how Philly is doing overall as a city, and how well its government is performing, boiled down to just two numbers.
Last updated 2026-07-30. Data through 2026-06.
Cite as: Philadelphia Performance Index (Nick Jain, last updated 2026-07-30). nickjain.com/resources/philadelphia-indexes
Performance Index - Latest
52.0*
2026-06 · -7.0 MoM
Partial — one or more contributing legs are current best estimates.
Government Index - Latest
53.3*
2026-06 · +0.2 MoM
Partial — one or more contributing legs are current best estimates.
How to read these scores. Each subscore and composite is normalized to 0–100 where 50 = typical. As a reading rule: ±10 = a normal swing, ±20 = notably good or bad, ±30+ = a structural shift or one-off event. Higher means better performance for every metric.
Monthly index and trailing 12-month average · As of June 2026
Monthly index and trailing 12-month average · As of June 2026
Each monthly sub-score reflects the latest available source data, including held-forward workforce and fiscal values. Asterisks mark partial-cohort or held-forward legs.
Each monthly sub-metric is scored against a baseline where 50 = typical. One robust standard deviation of relative deviation from baseline moves the score by 10 points, so a tight-variation metric and a wide-variation one both read on the same ±10 band. The scale uses the median absolute deviation (robust to COVID-era outliers) rather than the raw standard deviation. Scores are clipped to 0-100; higher is better. The composite chart's y-axis is zoomed to 25-75 so month-to-month movement is visible. See the May 2025 launch post for the origin story. Every monthly composite rebuilds against the data available today — including held-forward legs from quarterly sources — so a row's value is current-vintage. The asterisk on a composite or leg marks partial-cohort or held-forward inputs; both will revise as later data lands.
Baseline = 2022-2024 calibration window. Every metric uses 2022-2024 as its calibration period; the score's scale is the MAD of the relative-deviation series over those 36 months, so each subscore averages exactly 50 across the calibration window by construction. Monthly metrics with seasonality use per-month baselines (12 individual values across 2022/2023/2024); metrics with weak seasonality use a single 36-month mean; quarterly metrics use per-calendar-quarter baselines (4 values). Charts and tables show January 2020 onward — the performance index is bounded by the public Eviction Lab series (which begins in 2020 — pre-2020 monthly records are not available); the government index starts the chart at 2020 to share the performance index's x-axis.
Philadelphia Performance Index (6 legs, equal-weight): violent crime per 100,000 residents, fatal traffic crashes per month, a housing sub-composite (mean of eviction filings + property transfers per 100,000 + sheriff-deed median sale price), a jobs sub-composite (mean of the Philadelphia-Camden-Wilmington MSA unemployment rate and nonfarm payrolls), EPA monthly average AQI for Philadelphia County, and 311 service-request demand per 100,000 residents (excluding “Information Request” — fewer requests signal fewer things going wrong in residents' lives, so the leg is inverted).
Philly Government Performance Index (5 legs, equal-weight): service-balanced median 311 completion time per request month, share of L&I code violations resolved within 90 days of issuance, building permits + business licenses per calendar month, overtime ÷ base pay for salaried city employees (publishes per fiscal quarter; one quarterly observation populates all three months of its quarter, and months outside every published quarter score null), and fiscal health (QCMR revenue vs. budget + ACFR pension funded ratio).