A data-science study of how visible women's economic activity is: measurement model, uncertainty, sensitivity, scenarios and value of information - Samoa

The Gender Economic Visibility Index (GEVI) asks whether women’s economic activity is present, measured and recognised in the systems that set policy, budgets and taxes. This paper applies the index to Samoa as a data-science measurement study: 13 indicators, each converted to a parity score and graded for evidence quality, are combined into five dimensions and aggregated by a geometric mean into a score out of 100, which is then tested for uncertainty (10,000 joint Monte Carlo draws), sensitivity (seven weighting schemes and ten alternative specifications), attribution of the gap to full parity (Shapley values), peer rank, scenarios and a pathway to the 65-point Emerging threshold, and value of information for the data strategy.

Samoa scores 59.3 out of 100 (90% interval 55.1 to 63.3), in the Partial band, ranking 2nd of 15 economies (5th to 95th percentile of the rank: 1 to 3) and second highest in the series. Women in work do very well on job quality and firm ownership, so D2 Job quality and seniority (0.938) is the strongest dimension, but only 30.7% of women are in the labour force against 54.5% of men, and young women’s unemployment is 21.4% against 8.52% for young men. D4 Institutional visibility is the weakest dimension (0.396), while the largest share of the 40.7-point gap to full parity, 11.7 points (29%), is attributed to D5 Statistical and political voice. Of the 13 inputs, 12 are grade A and one, WBL enforcement perceptions, is imputed (grade C), giving a blind-spot ratio of 0.08 that widens the interval. The rank holds at 2 across all seven weighting schemes and all ten alternative specifications, and the score moves by at most 5.0 points. The largest single-lever scenario, halving the labour access gap to parity, is worth +4.4 points; a greedy pathway of +0.10 steps on measured indicators reaches the Emerging threshold after 6 steps, or about 4 years at an illustrative 0.02 a year. The findings bear directly on tax administration and fiscal transparency. All numbers are research estimates, not official statistics, and are fully reproducible from the accompanying scripts with a fixed random seed.

The paper is organised so that every number can be traced and challenged. It begins with the purpose and scope (Chapter 1), a data audit of sources, input ledger, data age and coverage (Chapter 2), and the measurement model (Chapter 3), followed by indicator-by-indicator construction (Chapter 4) and dimension scores with a full aggregation trace and Shapley attribution of the gap (Chapter 5). It then reports the uncertainty analysis (Chapter 6), sensitivity and robustness tests (Chapter 7), peer benchmarking and rank uncertainty (Chapter 8), and scenario simulation and pathways (Chapter 9), before turning to the value of information and a data roadmap (Chapter 10), findings, recommendations and a twelve-month implementation plan for Samoa (Chapter 11), and the limitations and validity threats of the approach (Chapter 12).