# Gaps & caveats — Productivity & informality SEA

**As of:** 2026-09-18 · pack `topics/productivity_informality/`

## Dual-story firewall

- **Labour productivity** (`gdp_per_employed_ppp`) and **informality** (`informal_employment_pct`) are **separate panels**. Never average, never fuse into one composite.
- **Self-employed %** (`self_employed_pct`, WDI `SL.EMP.SELF.ZS`) is a **secondary companion only**. It is **not** a substitute for ILO informal employment when the informal cell is empty.

## Informal employment — intentional empties

| ISO3 | Status | Note |
|------|--------|------|
| **PHL** | **GAP** | No `SDG_0831_SEX_ECO_RT` series in Asia ILOSTAT extract (train+test combined). Do not invent; do not backfill from PSA class-of-worker or self-employed %. |
| **MYS** | **GAP** | Same — absent from ILOSTAT SDG 8.3.1 Asia extract. |
| **SGP** | **GAP** | Common high-income coverage gap for informal employment SDG 8.3.1. |

## Informal employment — filled but caveated

| ISO3 | Latest | Caveat |
|------|--------|--------|
| **KHM** | 89.39% (2019) | `obs_status=B` (break / caution); sparse series (2012, 2019 only). |
| **LAO** | 90.49% (2022) | `obs_status=B`; sparse (2017, 2022). |
| **MMR** | 80.71% (2020) | Last ILOSTAT print **2020** — conflict / survey interruption caveat; do not plot as a 2024–25 print. |
| **TLS** | 80.57% (2021) | Small-economy LFS; badge year clearly. |
| **BRN** | 21.39% (2024) | Covered (contrary to older “often missing” note); still a tiny labour market — show separately from large AMS peers. |
| **IDN** | 80.97% (2023) | National LFS via ILOSTAT. |
| **THA** | 63.17% (2024) | Informal Employment Survey (HS) source label. |
| **VNM** | 67.03% (2024) | LFS via ILOSTAT; national GSO prints may differ slightly — lock to ILOSTAT SDG_0831. |

## Productivity caveats

- WDI `SL.GDP.PCAP.EM.KD` uses **ILO modelled employment** + PPP GDP (constant **2021** PPP $). Country comparisons with thin national employment data need caution (ILO modelled-estimates guidance).
- All SEA-11 have **2025** productivity observations in the WDI pull (API `lastupdated` **2026-07-13**). Badge **2025**.
- **Singapore** and **Brunei** dominate the level scale — treat as hub / resource outliers, not PH peers.

## Definition divergence

- Informal employment (ICLS / ILO) ≠ informal *sector* ≠ self-employment ≠ vulnerable employment.
- Self-employed % can move with agriculture share and own-account work without matching ILO informal %.

## Closed GAP (cross-pack)

- `topics/labour_youth/scoreboard.csv` documented an **informal employment GAP** on every row. This pack is the dedicated fill for that metric where ILOSTAT has data; PH/MY/SG remain empty here too (honest).