The revision tracker: first prints vs revised truth

Initial releases versus what the numbers became after revision — payrolls, CPI seasonal factors, GDP — with the size and direction of every revision on the record. Markets trade the first print; the economy runs on the revised one.

Live readings

• Nonfarm payrolls (current vintage): 159,044 k (2026-09-01, +0.3% YoY)

Payroll revisions run tens of thousands of jobs and turn systematically negative near turning points — a real-time recession tell. This page uses vintage data (what was published, when) — the same discipline behind this site's own vintage-true forecast backtests.

What this page measures

Every macro number is published twice or more: the first estimate the market trades on, and the revised value the economy actually ran on. This tracker reads the St. Louis Fed's ALFRED archive — every vintage of a series exactly as it stood on each publication date — and lines up, for 4 headline releases since 2015, what was first reported against what the number is today. "Initial" is the figure as first seen; for payrolls and retail sales it is computed from the exact vintage published on release day.

The record, release by release

The summary statistics cover the last 24 prints and exclude the newest one, which has not yet had a chance to be revised. A negative average revision means first prints have tended to overstate the number; the share revised down says how one-sided that has been.

Nonfarm Payrolls (monthly change) — last six prints

Retail Sales (MoM %) — last six prints

CPI (MoM %) — last six prints

Real GDP Growth (SAAR %) — last six prints

How to read revisions

Payroll revisions run to tens of thousands of jobs — the average over the last 24 prints is -58k, with 79% revised down and a cumulative -284k over the last twelve — and they turn systematically negative near turning points, because the birth–death model that estimates new firms keeps adding jobs after hiring has stopped. That makes the direction of revisions a real-time recession tell in its own right. CPI is barely revised in level terms (only seasonal factors are recalculated each February), which is why an inflation forecast can be graded against a first print without ambiguity. GDP moves the most: the advance estimate is built with incomplete trade and inventory data and is followed by second and third estimates and annual benchmarks. This is the same discipline behind Nowflation's own numbers: the Gauge's history is stored vintage-true and never rewritten, the scoreboard grades frozen rows, and the forecast backtest uses only the data that existed the day before each print.

Terms on this page

Vintage (vintage-true data)

A vintage is a statistic exactly as it stood on a given date, before later revisions; a vintage-true backtest uses only the data that existed the day before each release, so it cannot borrow information from the future. Nowflation stores its own history vintage-true and never silently revises a published value.

Seasonal adjustment (SA vs NSA)

Seasonal adjustment removes the regular calendar pattern (holiday sales, summer gasoline, January price resets) so month-to-month changes can be compared; BLS quotes the monthly change on the SA index and the year-over-year rate on the NSA index, where seasonality cancels. Nowflation follows the same convention, and its pace figures use the BLS gasoline factors.

ALFRED

The St. Louis Fed's archival database of FRED series by vintage — every value as it was published on each date.

Benchmark revision

The annual re-anchoring of the payroll survey to unemployment-insurance tax records, which can move a year of job gains by hundreds of thousands.

Birth–death model

The BLS adjustment that estimates jobs at businesses too new to be in the survey sample; it lags at turning points.

Related: The scoreboard · Jobs · Current US inflation rate · Methodology