MENA Open Data & Evidence Lab
MENA Evidence Note · Statistical Asynchrony 001 · 21 Aug 2026

The Economy Exists Before the Statistic Does

A first measurement of economic-information latency across MENA — and why the clock itself has to be defined before countries are compared.

Q1 2026 ended on 31 March. The economic quarter was the same calendar object across Saudi Arabia, Morocco, Jordan and Oman. The statistical information available about it was not.

29 → 103

days after quarter-end · recorded release objects in the current seed

EconomyRecorded release objectDateDays after Q1
Saudi ArabiaFlash GDP estimate29 Apr 202629
Saudi ArabiaFuller Q1 GDP release8 Jun 202669
MoroccoQ1 national accounts release29 Jun 202690
JordanObserved English Q1 GDP page5 Jul 202696
OmanQ1 national-accounts catalogue publication12 Jul 2026103
Important qualification. The 74-day spread from 29 to 103 days is a real spread across the recorded release objects, but it is not yet a harmonized estimate of “first public information.” Saudi Arabia publishes a flash estimate. Oman’s formal catalogue date is later than evidence of preliminary Q1 figures in public reporting. Jordan’s English page and archive metadata expose different July dates. MENA-RTD preserves these distinctions rather than choosing one date silently.

The result is the architecture

A normal historical panel downloaded today suppresses this problem. Once all values have arrived, every Q1 observation sits neatly next to every other Q1 observation. That tidy panel is not the information set that existed in real time.

This is what we call statistical asynchrony: at a given historical moment, economies are observable at different statistical depths. Release speed is only one dimension. Release architecture — flash estimate, preliminary release, full release, archive record, revision — is another.

A different dataset

MENA-RTD records releases rather than overwriting them. The key object is the historical information set: all source-verified observations that were public by date d. That makes it possible to reconstruct exactly what an analyst could have used on, for example, 15 May 2026, and prevents revised future information from leaking backward into pseudo-real-time analysis.

What comes next

The immediate build is a 2015–2026 release clock across ten economies and five indicator families: GDP, CPI, unemployment, industrial production and trade. The target is at least 5,000 source-linked historical releases. That is a target, not a current count.

Once the archive is dense enough, the first analyses are: median publication latency by country and indicator; release-latency distributions; regional information-completeness curves; T50/T75/T90 information thresholds; first-versus-later GDP estimates; news-versus-noise revision tests; and models of whether faster release is associated with larger subsequent revision.

Source record

Data: mena_release_clock.csv · Schema: mena-rtd-schema.json · Method: MENA_RTD_METHODS.md · Code: analyze-mena-release-clock.py