Papua New GuineaWork Without Visibility
Near-parity in participation. A deep gap in the institutions that make women’s work visible to policy.
Women take part in work at almost the same rate as men, but almost everything that would make that work visible to policy is missing. Papua New Guinea is the pilot economy for the index: all 13 inputs are measured statistics.
Source: the attached country paper. Animation reveals the reported measures; it does not show change over time.
Abstract
Papua New Guinea records a GEVI of 47.2, with a 90% interval of 41.4 to 53.0, placing it in the Hidden band and ranking it 13th of 15 economies. Its labour participation score of 0.912 contrasts sharply with an institutional visibility score of 0.172. All 13 inputs are measured statistics, giving a blind-spot ratio of 0.00.
Women’s vulnerable employment is 79.5%, compared with 63.5% for men, while wage and salaried employment is 14.5% for women and 27.6% for men. These gaps limit the extent to which women’s work appears in the formal economy observed by pay-as-you-earn systems. The weighted geometric mean penalises this imbalance: raising institutional visibility alone from 0.172 to 0.30 would increase GEVI from 47.2 to 51.3.
The paper prioritises sex-disaggregated taxpayer and employer records, an annual gender tax profile, and clearer separation between reforms a revenue authority can deliver and political representation decisions. As the pilot economy, Papua New Guinea provides a reference for annual replication of the same 13 calculations.
Download the paper PDFHeadline result
The strongest dimension is labour participation; the weakest is institutional visibility. The index exposes the imbalance between taking part in work and being visible within formal institutions.
90% interval 41.4–53.0.
Rank 13 of 15 · 13 measured inputs.
versus D1 participation of 0.912
Women take part in work at almost the same rate as men, but almost everything that would make that work visible to policy is missing.
How the index is built
Each dimension is the simple mean of its indicators. GEVI is then calculated as the weighted geometric mean of the five dimension scores.
| Dimension | Score D | Weight w | D^w |
|---|---|---|---|
| D1 Labour participation | 0.912 | 0.25 | 0.9772 |
| D2 Job quality and seniority | 0.480 | 0.20 | 0.8636 |
| D3 Enterprise and legal rights | 0.508 | 0.20 | 0.8734 |
| D4 Institutional visibility | 0.172 | 0.15 | 0.7679 |
| D5 Statistical and political voice | 0.402 | 0.20 | 0.8334 |
= 47.2
Product of the five weighted terms: 0.4717. Dimension values and the index are rounded as reported in the source.
Uncertainty and rank sensitivity
The 10,000-draw simulation samples each score using the grade A standard deviation of 0.02 and draws dimension weights from a Dirichlet distribution centred on the base weights. The median is 47.1 and the 5th–95th percentile interval is 41.4–53.0. Under 5,000 random weight sets alone, the rank has a median of 13 and a 5th–95th percentile range of 9–14.
Where the gap is
Raising each dimension to full parity, while holding the others fixed, gives the following scenarios. Institutional visibility offers the largest modelled gain.
| Dimension raised to 1.0 | GEVI | Gain |
|---|---|---|
| D1 | 48.3 | +1.1 |
| D2 | 54.6 | +7.5 |
| D3 | 54.0 | +6.8 |
| D4 | 61.4 | +14.3 |
| D5 | 56.6 | +9.4 |
Single-indicator scenarios
Each scenario lifts one indicator to the upper quartile of countries with data.
| Indicator lifted | From | To | GEVI | Gain |
|---|---|---|---|---|
| I11 Women ministers | 0.054 | 0.373 | 52.1 | +4.9 |
| I12 Women MPs | 0.054 | 0.382 | 50.5 | +3.3 |
| I6 Wage and salaried employment | 0.525 | 0.984 | 49.9 | +2.7 |
| I4 Non-vulnerable employment | 0.562 | 0.989 | 49.7 | +2.5 |
Scenario values and gains are reproduced as printed in the paper; rounding can mean a displayed gain differs from subtraction of the displayed index values. These are conditional model scenarios, not forecasts or causal estimates.
Work is already present
Labour force participation is 50.8% for women and 53.3% for men.
More vulnerable work
Women’s vulnerable employment is 79.5%, compared with 63.5% for men.
Largest parity scenario
Lifting D4 to full parity gives a GEVI of 61.4 in the paper’s scenario.
Priorities and practical use
The priorities connect the index to actions and identify where revenue administration can act.
Make formal women’s work countable. Add a sex field to employer salary and wages tax returns and to the taxpayer register. Report wage employment by sex using administrative data, addressing I6 and giving D4 a fiscal indicator.
Publish an annual gender tax profile. Report who files, who is withheld from and who receives refunds, feeding I10 and I13.
Recognise the boundary of tax administration. Women hold 2.7% of ministerial and parliamentary roles. Quotas or reserved seats are political decisions; I13, the data indicator, is where a tax authority can act.
Who uses the data, and for what decision?
| Data product | Primary user | Decision it informs | Cadence |
|---|---|---|---|
| I4 and I6: vulnerable and wage employment by sex | Revenue authority, Treasury | Where the wage-tax base is thin for women, and whether a simplified regime for informal earners is worth designing | Annual |
| I11 and I12: women in cabinet and parliament | Ministry responsible for women, UN Women, donors | Targeting of leadership support and the evidence base for the reserved-seat debate | After each election |
| I10 supportive frameworks (29%) and the D4 gap (+14.3 points) | Treasury budget office | Which missing mechanisms (budget tagging, data unit, equal-opportunity body) to fund first | Budget cycle |
| I13 coverage (9 of 12 rows) | National statistics office | Which three missing rows (account ownership, savings, unpaid care time) to add to the next survey | Survey plan |
Source and scope. This research summary reproduces the attached paper’s reported findings, inputs and scenarios. The uncertainty interval and weight sensitivity qualify the interpretation of the headline score. A zero blind-spot ratio means all 13 index inputs are measured; it does not mean all policy-relevant data exist. I13 covers 9 of 12 rows.
Series: GEVI Research Series · DOI: 10.5281/zenodo.23204585