Same Label, Different Statistic · Data Note 002 · 17 August 2026
Five countries publish “unemployment.” The metadata underneath the number is not equally visible.
Jordan, Egypt, Morocco, Saudi Arabia and the UAE all publish labour-market statistics. This note audits the measurement perimeter and the public documentation needed to compare them responsibly. It is a metadata-comparability study, not a ranking of labour-market performance.
Finding 01
A shared headline label does not produce a shared information set.
Jordan, Morocco and Saudi Arabia expose enough current public documentation to code age framework and population perimeter directly. Egypt exposes a documented civilian-labour-force perimeter in the survey metadata reviewed. The UAE publishes a federal labour-force series and a 2024 headline unemployment rate, but the public pages reviewed for this release did not expose denominator, age framework and population perimeter with the same precision. The correct coding is therefore “metadata unresolved,” not an inferred definition.
Finding 02
Morocco and Saudi Arabia show why a time series needs a methodology flag.
Morocco · explicit redesignEMO2026 is a new generation of labour-force survey.
HCP describes the 2026 survey as a new-generation labour-force survey aligned with recent international labour-statistics standards. It explicitly separates the labour force from the population outside it and identifies a potential labour force outside strict unemployment.
Research consequence: a pre/post-2026 comparison should carry a method-change flag rather than silently chaining the headline rate.
Saudi Arabia · explicit redesignThe Saudi LFS questionnaire changed from Q1 2025.
GASTAT documents a redesigned questionnaire from Q1 2025, adoption of ICSE-18, continuous weekly sampling, and a 15+ Saudi/non-Saudi resident target population. It also documents comparable time-series groupings by census reference.
Research consequence: the series can be used, but the methodological transition belongs in the observation metadata.
| Country | 15+ resolved? | Population perimeter | Method-change flag | Public metadata state |
|---|
| Jordan | Yes | Releases distinguish Jordanian and non-Jordanian populations | No 2026 redesign flagged here | Resolved |
| Egypt | Yes in reviewed survey metadata | Egyptian civilian labour force inside the State | Not assessed as a 2026 redesign | Documented perimeter |
| Morocco | Yes | Labour force vs outside-labour-force architecture; potential labour force explicit | EMO2026 new-generation survey | High detail |
| Saudi Arabia | Yes | Saudi + non-Saudi residents in private housing, all regions | Q1 2025 questionnaire redesign | High detail |
| UAE | Not resolved from reviewed pages | Federal labour-force series exists; exact denominator/perimeter not coded here | No redesign conclusion made | Partial |
Method
We coded only fields that were visible in official public sources reviewed on 17 August 2026. A missing definition was retained as missing. We did not backfill a national definition from third-party databases, and we did not treat a published unemployment rate as proof that denominator and population perimeter were publicly documented on the same page.
Interpretation boundary
This note does not claim that one country measures unemployment “better.” It measures whether the information needed for cross-country interpretation was visible and specific in the official public material reviewed. Public metadata can change; future releases should version this matrix.
Suggested citation: MENA Open Data & Evidence Lab (2026). Same Label, Different Statistic: Five-country unemployment comparability audit. Data Note 002, version 1.0.