Accountability · the ledger

Claim check

This ledger checks public claims against independent, matched-period data — regardless of who made them. Where a claim is directionally true but exaggerated, we say so; where a critical reading is itself overstated, we say that too. When figures genuinely can't be reconciled, the verdict is Unverifiable, not a guess.

Holds upNeeds contextOverstatedMisleadingUnverifiable

MisleadingSafety· Late 2025

“Street crime in the province fell by 70–80% in 2025.”

— Provincial government (Punjab)

The verdict

Directionally real, wildly overstated — the matched-period change was closer to ~2%.

The claim was checked directly against the province's own police data on a like-for-like, matched-period basis. Crime did edge down, but by roughly 2% — not 70–80%. A gap of that size between the headline and the underlying data is exactly why this dashboard treats government crime-reduction claims as unproven until verified against multi-year, matched-period figures.

Checked against

OverstatedEconomy· Mid–late 2025

“Inflation has been brought durably under control.”

— Official economic messaging (2025)

The verdict

The 2025 low was real but temporary; inflation was back in double digits within a year.

Headline CPI genuinely fell to multi-year lows in 2025 (about 3.1% in August 2025), and that disinflation was real. But framing it as durable control was premature: much of the dip was a high base effect and softer commodity prices, and by June 2026 inflation had re-accelerated to 11.1% — above the SBP's 5–7% target band. 'Eased for now' held up; 'durably controlled' did not.

Checked against

MisleadingSafety· 2025–26

“The security situation is steadily improving.”

— Government security messaging

The verdict

Independent trackers record 2025 as the deadliest year in roughly a decade.

Against the national trend, this doesn't hold: independent trackers agree 2025 was the most violent year in about ten years — PIPS logged 699 attacks (1,034 fatalities, +34% YoY), and broader combat-death counts rose similarly. The improvement narrative only survives at the level of specific metros; nationally, and especially in KP and Balochistan (96% of fatalities), violence intensified.

Checked against

UnverifiableEconomy· Jul 2026

“Foreign-exchange reserves have been rebuilt to about $25 billion.”

— S&P rating rationale, echoed in official messaging

The verdict

Reserves are genuinely rebuilding, but the $25.3bn figure and SBP's own $17.2bn haven't been reconciled.

Reserves have clearly recovered from crisis lows — that part is not in dispute, and it underpinned an S&P upgrade to 'B'. But the widely-cited $25.3bn (end-June 2026, per S&P) sits well above the State Bank's own weekly-published SBP-held figure of ~$17.2bn (10 Jul 2026). The two almost certainly measure different aggregates (gross vs. liquid, or different dates), but the exact reconciliation is not public — so the specific '$25 billion' headline is, for now, unverifiable rather than confirmed.

Checked against

Needs contextEconomy· 2026

“The jump in the poverty rate to ~48% shows a pure, dramatic real-world collapse.”

— Common public reading of the World Bank figure

The verdict

Poverty is genuinely high, but part of the rise from 44.7% is the World Bank raising the line, not only real deterioration.

This one cuts against the alarmist reading, not the government. Measured poverty did rise to 47.9%, and hardship is real. But a meaningful part of the increase from the earlier 44.7% reflects the World Bank rebasing its international poverty line upward (to $4.20/day, 2017 PPP) — a definitional change, not solely households falling below a fixed bar. Read the level as serious; read the *jump* as partly methodological.

Checked against

The ledger grows as new claims are made and checked. Every verdict rests on independent, matched-period data carried elsewhere in this snapshot — click any “checked against” line to interrogate the evidence. Nothing here is adjudicated beyond what the data can support.