MacroDataHub

Tracking the US economy weekly

GDP arrives a month after the quarter ends. Four indicators try to fill the gap with weekly or daily readings, and each buys its timeliness with a different compromise.

At a glance

MeasureLatestPeriodPublished byFrequency
Weekly Economic Index (WEI)3.03%Sep 19, 2026Federal Reserve Bank of DallasWeekly
ADS Business Conditions Index−0 Standard deviationsSep 19, 2026Federal Reserve Bank of PhiladelphiaDaily
CARTS weekly retail sales629,178.7 USD millionAug 28, 2026US Census BureauWeekly
Initial jobless claims197,000 PersonsSep 19, 2026Department of LaborWeekly

How each one is built

Weekly Economic Index (WEI)
Method: Ten weekly series scaled to match four-quarter GDP growth
Covers: Retail, labour and production indicators; reads as an annualised growth rate
Released: Thursday
ADS Business Conditions Index
Method: Mixed-frequency model blending daily, weekly, monthly and quarterly inputs
Covers: Zero = average conditions; revised continuously as data arrive
Released: Updated as inputs arrive
CARTS weekly retail sales
Method: Advance estimate of retail and food services sales from third-party transaction data
Covers: Retail only, national
Released: Within about two weeks
Initial jobless claims
Method: Administrative count of new unemployment insurance filings
Covers: The oldest and cleanest weekly series — a count, not an estimate
Released: Thursday, for the week ending Saturday

What you give up for speed

The WEI and the ADS index are model constructs. They combine series that were never designed to be read together, and they revise — the ADS index in particular is rewritten every time a new monthly input lands, so the value you quoted last week may no longer exist. Anyone citing a level must say which vintage they used.

Initial claims revise far less, because they are a count of filings rather than an estimate. Their weakness is the opposite: they are noisy week to week and are distorted by holidays, seasonal factors and state-level administrative changes. The four-week average exists for exactly that reason.

Which one to use

For a single number that maps onto GDP growth, use the WEI — it is scaled so that a reading of 2 corresponds roughly to 2% four-quarter GDP growth, which no other weekly series offers.

For turning points, watch initial claims and the four-week average. They are the earliest hard data on the labour market, they do not depend on anyone’s model, and every US recession has been preceded by a sustained rise in them.

Use CARTS for consumer spending specifically, and remember it covers retail alone — services, which are most of US consumption, are not in it.

Frequently asked questions

How should I read the Weekly Economic Index?
It is scaled to four-quarter GDP growth, so a value of 2 corresponds roughly to 2% year-on-year GDP growth. It is not a level of output.
Why does the ADS index keep changing?
It is a mixed-frequency model that re-estimates the whole history whenever a new input arrives, so past values are revised continuously.
Are initial jobless claims revised?
The prior week is revised once, usually by a small amount, when states report final counts. Older weeks are stable.
What does CARTS not cover?
Services spending. It covers retail and food services only, which is under half of US personal consumption.
MacroDataHub on Facebook