Digital Skills Gap Tracker
Project 6 · Georgetown University Partnership · Spring 2026
DEMO
Global Digital Skills Overview
Market Demand
Intervention Simulator
Gender Gap
How To Guide

Purpose

Static reports often fall short when communicating complex digital skills insights to policymakers and senior leadership. This interactive tool lets users explore the data themselves — filtering by country, skill, and gender — making the message land with greater clarity and impact.

Case in point · Ethiopia

We'll use Ethiopia (pop. ~126M, Sub-Saharan Africa) as the recurring example through this dashboard — current state on the map, gender gaps on the Gender tab, demand pull on Market Demand, and a concrete program sequence on the Simulator.

Basic Lit.20
Data / Analytics65
Cybersecurity18
AI / ML38
Cloud25
Foundational layers (basic literacy, cybersecurity, cloud) are the binding constraints. Data & Analytics is a relative strength. Scores 0–100 from 5 independent indicators — WDI internet use, WDI scientific articles, ITU GCI 2024 tier, Oxford AI Readiness 2024, WDI secure servers.
12/20
Weakest digital skills countries are in Sub-Saharan Africa
▼ Skills gap widening with AI adoption surge
11%
Africa tertiary graduates with formal digital training
▼ Far below 38% global average
15%
Gender gap in mobile internet use (women vs. men)
▼ Persistent across all income groups
78%
Businesses now using AI — up from 55% in 2023
▲ Demand for digital talent accelerating →
Sources for key statistics: World Bank Digital Progress & Trends Report 2025 · Stanford AI Index 2025 · World Bank Gender & Digital Inclusion
1 Digital Skills Proficiency Map — 30 Developing Countries ▶
Countries are colored by average digital skills score (0–100). Click any highlighted country to see a deep dive with proficiency tabs across all 5 skill domains.
Avg score:
<2020–3435–4950–6465+ Not in dataset
Data sources: World Bank WDI (D1 IT.NET.USER.ZS · D2 IP.JRN.ARTC.SC · D5 IT.NET.SECR.P6) · ITU GCI 2024 (D3 cybersecurity tier) · Oxford Insights AI Readiness 2024 (D4 AI/ML Total Score) · ✔ All 5 domains use independent real indicators (no derivation across columns). See methodology below.
How country scores are constructed

Each country receives a score from 0–100 for each of the 5 skill domains. Every domain is now anchored to an independent public indicator (no within-country derivation across columns), so cross-country comparison within a domain and cross-domain reading within a country are both meaningful. Sources:

  1. Basic Digital Literacy — WDI IT.NET.USER.ZS (individuals using the Internet, % of population). Linearly scaled (×0.90) to 0–100. Source: World Bank WDI, most recent year per country (2022–24).
  2. Data & Analytics — WDI IP.JRN.ARTC.SC (scientific & technical journal articles) per million population, log10-normalized (anchors: 2 articles/M → 0, 316 articles/M → 100). Proxies the analytical/quantitative research workforce — a structural pre-condition for downstream data & analytics capacity. Source: World Bank WDI (NSF / Scopus, 2022).
  3. Cybersecurity — ITU Global Cybersecurity Index 2024 tier classification (T1=87, T2=63, T3=40, T4=18). Source: ITU GCI 2024 (Sept 2024 release).
  4. AI / ML — Oxford Insights Government AI Readiness Index 2024 Total Score (0–100, equally-weighted mean of Government, Technology Sector, and Data & Infrastructure pillars across 39 indicators). Source: Oxford Insights, Dec 2024 edition. Tanzania & Guinea-Bissau use SSA-regional median (≈42.6, rounds to 43) as fallback.
  5. Cloud Computing — WDI IT.NET.SECR.P6 (secure Internet servers per million people), log10-normalized (anchors: 3 servers/M → 0, 10,000 servers/M → 100). Direct measure of TLS-secured server infrastructure — the lower bound on cloud-deployable capacity in country. Source: World Bank WDI / Netcraft (2024).
D1 = min(internet × 0.90, 100)  |  D2 = clamp((log10(art/M) − 0.3) / 2.2 × 100, 0, 100)
D3 = TIER_SCORE[ITU GCI tier]  |  D4 = round(Oxford Total Score)
D5 = clamp((log10(servers/M) − 0.5) / 3.5 × 100, 0, 100)

Data status: All 5 domains now use real, independently-fetched indicators (no derivation across columns). The data_refresh.py script pulls D1, D2 and D5 from the World Bank API and embeds the hardcoded ITU GCI 2024 tier table (D3) and Oxford Insights 2024 Total Score table (D4). Two countries fall back to regional medians where source coverage is missing: Sudan & Myanmar (WDI internet — D1 only); Tanzania & Guinea-Bissau (Oxford 2024 — D4 only). All other (country × domain) cells are real measurements.

2 Regional Skills — Two Lenses ▶
The same 30 countries × 5 skill domains, viewed two ways. Compare regional skill profiles head-to-head, then drop down to country-level distribution. Both views are clickable for drill-down.
Source: World Bank Data360 · UNCTAD Digital Economy Reports

Why this matters: the size of the opportunity

The supply-side picture (skills proficiency, gender gaps) only tells half the story. For investors and donors weighing whether to fund digital-skills interventions, the demand side answers a sharper question: how big is the market employers are trying to fill, and what will trained workers earn? The figures below are drawn from the World Bank's Digital Progress and Trends Report 2025 — page numbers cited throughout.

Make this tab about your market
Personalize for a country to see how the global numbers translate locally.
1 Demand-side fundamentals — at a glance ▶
Five headline figures from WB DPTR 2025, grouped by what they tell an investor: how big the market is and how fast it's growing, what trained workers earn, and what employers are actually doing when they hire.
Market size & growth
1.7M
Global annual AI job vacancies (≈1.5% of all online vacancies, 2021–2024)
▲ Steady demand for AI talent worldwide · Source: WB DPTR 2025, p. 78
9×
Surge in GenAI job vacancies globally, 2021 → 2024
▲ The fastest-growing skills category · Source: WB DPTR 2025, p. 79
Employer behavior
60%
IT-services job postings that require advanced digital skills
▲ 53% in pro services · 44% in finance · Source: WB DPTR 2025, p. 76
75%
Employers preferring less-experienced candidates with GenAI skills over more-experienced ones without
▲ Skills now outweigh seniority for hiring · Source: Coursera 2025 Survey via WB DPTR 2025, p. 83
Worker payoff
25–36%
Wage premium for GenAI literacy in non-tech white-collar roles
▲ Direct ROI to trained workers · Source: Martins-Neto et al. (2025) via WB DPTR 2025, p. 79
2 2024 AI-vacancy snapshot — by dashboard region ▶
All 5 regions on one card, each shown in its native metric: 3 regions report stated 2024 absolute vacancy counts (left); the other 2 report only growth multiples 2021 → 2024 (right). Side-by-side avoids the apples-to-oranges problem of mixing counts and multipliers on the same axis.
Source: World Bank — Digital Progress and Trends Report 2025, p. 78 (Lightcast data)
3 Demand by industry & occupation — share of 2024 postings requiring digital / AI skills ▶
Each row in the WB DPTR 2025 occupation/industry tables is plotted as a point on the matrix below: the X-axis is the share of 2024 postings requiring advanced digital skills, the Y-axis is the share requiring AI skills (capped at 15% — no industry yet exceeds it). Quadrants are split by the WB-published 21% high-skilled global baseline (p. 76) on X and a 5% "notable AI demand" threshold on Y (derived heuristic — no WB-published Y baseline; see legend), translating each region of the chart into a training-strategy posture. Filled circles are rows where both metrics are published; hollow dashed circles on an axis mark rows the report publishes on only one metric. Click any point — or any chip in the off-chart strip — for a policy implication grounded in the report.
Show:
Source: World Bank — Digital Progress and Trends Report 2025, Ch. 5, pp. 76–78
4 The investor case — at a glance ▶
Four story cards: where the demand lives, the wage payoff for trained workers, the supply shortfall, and where to act next.
① Where the demand lives
~1.7M
global AI job vacancies in 2024
of which ~511K (30%) are in the 30-country tracker
0↑ click any segment1.7M global
📍 Anchor markets (2024)
~230K
South Asia
India-led
~225K
East Asia & Pacific
China-led
~56K
Latin America
Brazil-led
📈 Where it's accelerating fastest (documented growth multiples, no published absolute)
3–4×Sub-Saharan AfricaKenya & Nigeria, 2021→2024
2×M. East & N. AfricaEgypt, 2021→2024
Source: WB DPTR 2025, p. 78. → Open the full Market Demand tab for the industry & sectoral breakdown.
② Wage premium for trained workers
GenAI literacy lifts pay +25–36% in non-tech roles — the income channel that recovers training cost.
Non-technical white-collaranalysts, marketers, consultants
+25–36%
Technical rolesengineers, developers, data scientists
+7–9%
📐 Estimate the dollar uplift in your market
Bars scaled to a 0–46% range so the relative size of the two premiums is visible. Calculator uses 30.5% (mid-range) × that country's GDP-per-capita (WB WDI NY.GDP.PCAP.CD). Source: Martins-Neto et al. (2025), via WB DPTR 2025, p. 79.
③ Supply isn't catching up on its own
Only 5% of low-income populations have basic digital skills — vs. 66% in high-income countries.
5%
Low-income
countries
66%
High-income
countries
40% of African universities haven't updated their ICT curricula in 5+ years — supply can't scale on its own.
Source: WB DPTR 2025, p. 71 (basic skills), p. 82 (gap), p. 83 (curricula).
④ How interventions translate
Closing the gap → trained workers → demand met. Try it on the next tab.
Investment
Pick a budget
Trained workers
Cost-per-trainee model
Demand met
vs. regional anchor
The Simulator estimates how many workers a given investment trains and compares the output to the documented regional AI-vacancy anchor for the country you select.

What this tab tracks: the digital gender gap, end to end

Women in developing countries face compounding gaps — fewer phones and less daily internet use, lower enrollment in formal digital training, and dramatically lower representation in the AI, Data, Cloud, and engineering workforce. This tab pulls those gaps into a single view across three layers — access → skills → workforce — so policymakers and donors can see where each leak starts and design interventions that close them in sequence, not in isolation. Source data: World Bank Gender & Inclusion in Digital Development, ITU, and Stanford AI Index 2025.

Investor topline · 3 things to leave this tab knowing
1
Returns concentrate where the gap is widest.
Women are 14% of cloud and 32% of AI / data — the same roles paying a +25–36% wage premium.
2
Foundational access gates upper-skill ROI.
Ethiopia: women trail men by 29 pts on mobile ownership. The fix: subsidized smartphone + mobile-data bundles for women, paired with the GSMA Connected Women model that drove Rwanda's faster gap-decline.
3
The market isn't monolithic.
Vietnam is at near-parity (98% / 97% mobile ownership). Allocate by readiness: upskilling in parity markets, access first in gap markets.
32%
Women in Data / AI workforce globally
20%
Women in engineering roles globally
14%
Women in cloud computing globally
15%
Mobile internet gender gap (women less likely to use)
Sources for key statistics: World Bank Gender & Inclusion in Digital Development · World Bank Digital Progress & Trends Report 2025
1 Mobile Access Gaps by Country — ownership, smartphone, daily use & texting (women vs. men) ▶
Why access matters as much as skills. Understanding the gender gap in digital skills is essential — but it only tells half the story. The other half is the gap in access to the tools themselves: a phone, a smartphone, the daily connectivity, and the literacy to text. You cannot close a skills gap for people who do not yet have the device, the data plan, or the basic ability to use it. The charts below quantify that upstream gap so it can be addressed alongside training programs, not after them.
Mobile Access by Gender — compare by country or region
Each line shows the women-to-men gap. Countries are ordered by gap size — the longer the bar, the more women are excluded.
Source: World Bank Global Findex 2025 (gender-disaggregated, age 15+). Five countries (Rwanda, Haiti, Sudan, Guinea-Bissau, Myanmar) not yet covered for access metrics.
2 Digital Skills Parity by Country — women-per-10-men pictogram across skill domains ▶
Digital Skills by Gender — compare by country or region
Choose a scope and skill domain. Each row shows 10 avatars; women's icons are filled proportional to their score vs. men's. Click any country card for a deep dive.
Source: GSMA Mobile Gender Gap Report 2024 · World Bank Gender & Digital Inclusion · ✔ Parity ratio = Female score ÷ Male score, rendered as filled icons out of 10.
How gender scores are constructed

Each country's male and female digital skills scores are computed separately using gender-disaggregated indicators:

  • Mobile internet use by gender — GSMA Mobile Gender Gap Report + WDI (IT.NET.USER.ZS disaggregated where available)
  • Female tertiary enrollment in STEM — WDI: SE.ENR.TERT.FM.ZS (school enrollment, tertiary, female)
  • Female workforce in ICT roles — ILO ILOSTAT gender-disaggregated employment by sector (available for ~60 countries)
  • Mobile money and financial digital access — World Bank Findex gender data as a basic digital participation proxy
Female Score = 0.35×mobile_internet_use_female + 0.30×STEM_enrollment_female + 0.25×ICT_employment_female + 0.10×digital_finance_female
Gender Gap = Male Score − Female Score

Data status: Male scores are derived from real WDI internet usage data (via data_refresh.py). Female scores apply GSMA 2024 regional gender gap ratios — the best available free proxy, covering all 30 countries.

3 Trend Over Time — gender gap progress 2020→2025 ▶
Gender Gap Trend — Progress 2020→2025
Click country chips to add or remove lines (up to 8 at a time)
Source: World Bank Digital Jobs & Skills Brief · World Bank Gender & Digital Inclusion · ✔ Trend lines anchored to GSMA SSA regional gap trajectory (2020–2024). See methodology below.
How gender gap trends are constructed

The trend lines show how the gender digital skills gap (Male Score − Female Score) has changed from 2020 to 2025. In the final tool, this will be computed from:

  • WDI time-series — Annual female tertiary STEM enrollment (SE.ENR.TERT.FM.ZS), available via WDI Databank API
  • GSMA Mobile Gender Gap Reports 2020–2024 — Annual country-level female mobile internet adoption (downloadable from GSMA Connected Women)
  • ILO ILOSTAT — Year-by-year female ICT employment share, where available (ilostat.ilo.org)

Data status: Trend values are anchored to GSMA's published SSA regional trajectory: the Sub-Saharan Africa mobile internet gender gap narrowed from approximately 22% in 2020 to 15% in 2023, as reported across the GSMA Mobile Gender Gap Reports 2020–2024. Country lines are offset from this regional baseline using known policy context — Rwanda shows a steeper decline reflecting its Digital Ambition 2020 programme and GSMA Connected Women partnership; Nigeria tracks above the regional average reflecting a persistently larger gap documented in GSMA country data. The 2025 value is a one-year forward projection of the 2024 trend. Country-specific year-by-year data can be extracted from the GSMA annual report Excel downloads.

What this tab does: turn investment dollars into projected skills gains

Pick a country, an investment level, and one or more intervention programs (bootcamps, university STEM, online platforms, public-private partnerships, upskilling) — the simulator projects the 5-year change in each of the five digital-skill domains, the cost-per-skill-point, and the workers reached per $1M. Pin scenarios to compare side-by-side, save named scenarios for later, or use the five-program comparison card below to see which intervention gives the most lift per dollar in each domain.

1 Upskilling Intervention Impact Simulator ▶
Model how different investments affect national digital skills indicators over 5 years — based on World Bank intervention data
Pinned: 0 / 3 scenarios
Methodology sources: Martins-Neto et al. (2025) — "Click, Code, Earn" · World Bank Digital Jobs & Skills Brief · World Bank DPTR 2025 · ℹ This is a scenario exploration tool, not a predictive model — parameters grounded in World Bank literature. See methodology below.
How intervention projections are modelled

The simulator estimates how a given investment in a specific intervention type will shift a country's digital skills scores over time. The model structure is:

Projected Score (Year Y) = Baseline Score + (Investment_$M × Effectiveness × Y/5)
Capped at 100; annual gain distributed linearly across projection period

Effectiveness multipliers by intervention type — these represent estimated score-points gained per $1M invested per skill domain, currently set as illustrative values inspired by the literature:

  • Coding Bootcamps — Higher impact on Data/Analytics and AI/ML (short cycle, targeted); lower impact on foundational literacy. Real calibration source: World Bank Nigeria skills programme evaluation (3M trained)
  • University STEM Programs — More balanced across domains, but higher cost per point and longer time lag. Source: WDI STEM enrollment trends vs. workforce outcomes
  • Online Platform Rollout — Highest reach and lowest cost, but concentrated in Basic Literacy; limited impact on advanced skills. Source: Digital Economy for Africa initiative reports
  • Public-Private Partnership — Moderate effectiveness across all domains; strong on cybersecurity and cloud due to industry involvement. Source: World Bank Digital Jobs Brief PPP case studies
  • Upskilling Training — Workplace-based reskilling for workers already in the labor force. Highest effectiveness on AI/ML, Data/Analytics, Cybersecurity, and Cloud — the IMF documents that AI-developer new skills carry an ~8% wage premium in the US and that IT, business/data-analysis, and engineering skills lead all new-skill wage premiums. Basic Literacy stays low because adult upskilling targets workers who are already literate. Per-worker cost is higher than bootcamps (release time, tailored curricula), so reach is ~1,500 workers per $1M. Calibration source: IMF SDN/2026/001 "Bridging Skill Gaps for the Future: New Jobs Creation in the AI Age" (Jaumotte, Kim, Koll, Li, Li, Melina, Song, Tavares, January 2026), Chapter III — a 1pp rise in new-skill demand is associated with +2.3% wages and +1.3% employment in US local labor markets and +0.9% wages in Germany; retraining intensity moderates employment effects.

Workers reached — estimated from typical programme throughput rates (bootcamp: ~2,000/M; university: ~500/M; online: ~8,000/M; PPP: ~3,000/M; upskilling: ~1,500/M). Based on World Bank programme documentation and IMF SDN/2026/001 Ch. III for the upskilling row.

Multi-program portfolios — when more than one intervention type is checked, the total investment is split equally across the selected programs (per-program budget = total ÷ N), per-domain score gains are summed across programs (capped at +30 score-points per domain), and workers reached is the sum across programs. Combining complementary programs — e.g. Coding Bootcamps (strong on Data/Analytics & AI/ML) plus Online Platform Rollout (strong on Basic Literacy, very high reach) — typically produces a more balanced score profile than concentrating the same budget in a single intervention type, at the cost of lower depth in any one domain.

Gender gap reduction — modelled as 35% of total skills gain flowing to female participants, based on World Bank Gender & Digital Inclusion research showing women's disproportionate benefit from targeted programmes when access barriers are addressed.

Design intent: The World Bank project brief explicitly calls for a tool to "model how different upskilling investments would affect national skills indicators over 5 years" — framing this as scenario exploration, not prediction. The effectiveness multipliers are informed estimates grounded in the literature above, intended to illustrate relative trade-offs between intervention types rather than produce precise forecasts. This is consistent with how the World Bank uses scenario tools internally for investment prioritisation discussions. Empirical calibration using Martins-Neto et al. (2025) regression outputs would be a natural refinement in the final build.

2 Compare the Five Intervention Programs ▶
All five programs at a glance — ranked on five criteria designed to surface the cost / reach / depth trade-offs an investor or policymaker faces when picking a portfolio. Lower cost-per-point and higher reach-per-$1M are better; the depth profile shows where each program's spend goes. Note: the cost / reach figures below are catalog rates (program parameters, country-agnostic). The Simulator above scales them by country baseline (diminishing returns), foundational-literacy floor (advanced skills), and GDP-of-delivery (cost) — see the "Why these numbers?" panel under each scenario for the country-specific multipliers.
1Cost efficiency — $M needed to lift one score-point in a target domain. Lower is better.
2Scale potential — workers reached per $1M invested. Higher is better.
3Skill-depth profile — score-points gained per $1M, shown across all 5 domains so you can see where the spend lands.
4Top strength — the single domain the program is best at building.
5Composite ROI — average effectiveness ÷ cost-per-point: a single number summarising "score gained per dollar spent across all 5 domains."
How to read this: The bars are length-normalised within the comparison set — the longest bar = best on that criterion across all 5 programs. Composite ROI is a quick screen, not a portfolio recommendation: combining complementary programs (e.g. Online Platform + Coding Bootcamps) typically beats picking the single highest-ROI program alone, because no single intervention covers every skill domain. Use the simulator above to size and combine programs for a specific country.

About This Tool

Why interactive? Complex digital development insights are often difficult to convey through static reports alone — especially when communicating to senior leadership and government officials who need to act quickly. This dashboard enables users to explore, filter, and simulate — transforming data into decisions.

This tool is Project 6 of the Georgetown University–World Bank Digital Development Partnership (Spring 2026 Data Science & Public Policy Practicum). It visualizes digital skills workforce readiness and gender-disaggregated skills gaps across 30+ developing countries, serving as a live decision-support tool for the World Bank Digital Development Team.

The dashboard's 5 country skill-domain scores are anchored to independent real public indicators: World Bank WDI (Internet usage IT.NET.USER.ZS; scientific articles IP.JRN.ARTC.SC per million pop; secure servers IT.NET.SECR.P6 per million pop), ITU Global Cybersecurity Index 2024 tiers, and Oxford Insights Government AI Readiness Index 2024 Total Score. Demand-side context, wage premium and qualitative narrative draw on the World Bank Digital Progress and Trends Report 2025, GSMA Mobile Gender Gap Report 2024, and UNCTAD Digital Economy Reports.

How To Use The Tracker

This dashboard isn't a static report — it's a decision-support surface designed to be picked up mid-meeting, filtered to a country or region, and used to brief a partner in 5–10 minutes. The guidance below maps each tab to the question it best answers, then lays out concrete playbooks for the four audience types most likely to act on the data.

How to read the four tabs

Audience playbooks — who should use this and how

International Development Organizations

World Bank, regional development banks, UN agencies, bilateral donors

  • Use the Overview heatmap to set country prioritization for the next funding cycle — cross-reference with income group and existing portfolio coverage.
  • Use the Simulator's "Compare the Five Programs" panel to defend cost-per-skill-point trade-offs in board materials.
  • Use the Gender tab's mobile-ownership gap data to flag access-layer prerequisites before approving upper-skill projects — a bootcamp loan in a country with a 29-pt mobile gap underperforms by design.
Concrete action: Pair concessional lending for digital infrastructure (mobile, broadband, digital ID) with grant-funded skilling — sequence the foundational layer 12–18 months ahead of upper-skill programs.
Country Governments & Line Ministries

Ministries of Education, ICT / Digital, Labour, Planning & Finance

  • Identify your country's two weakest domains on the Overview heatmap — these are your binding constraints, not the headline AI/ML score.
  • Use the Simulator to size a national digital skills strategy: pick one foundational program (Online Platform or Upskilling) plus one upper-skill program (Bootcamps or PPP), then iterate on the $M and year inputs until the projected 5-year profile matches strategy targets.
  • Use the Gender Trend chart to benchmark your trajectory against regional peers — the underlying Rwanda policy package (Digital Ambition 2020 + GSMA Connected Women partnership) is documented in the dashboard's trend-methodology notes and worth studying as a model, even though the modelled decline trajectory itself is illustrative.
Concrete action: Anchor your national digital skills strategy in a 5-year sequenced portfolio, not a single flagship project. Reassess against the dashboard yearly as new WDI / ITU / Oxford data lands.
Investors & Private Capital

Impact funds, ed-tech, BPO operators, fintech, AI-services exporters

  • Use the Market Demand tab's sectoral mix to size the addressable workforce in your target geography — financial services, professional services, and ICT lead AI-vacancy share.
  • Cross-reference the Overview map to find geographies where supply gap (low D2/D4/D5) meets demand pull (high vacancy growth) — that's where pricing power on talent is greatest.
  • Use the wage-premium data (+25–36% for GenAI literacy) as the headline ROI figure when raising thesis-driven funds focused on workforce-readiness products.
Concrete action: Stage capital allocation by skill layer — seed / early-stage in foundational ed-tech (literacy, mobile-first), growth-stage in upper-skill BPO and AI-services export plays.
Researchers, Analysts & Practitioners

Academics, think tanks, consultancies, embedded analysts at the agencies above

  • All raw indicators are public and cited — see the Key Citations section below for direct PDF / API links to WDI, ITU GCI, Oxford GARI, GSMA, and the World Bank DPTR 2025.
  • The composite scoring methodology (5 domains × 5 independent indicators, normalized 0–100) is reproducible via data_refresh.py.
  • Use the dashboard as a briefing layer on top of your own deeper country analysis — the simulator is calibrated to IMF SDN/2026/001 and World Bank intervention data, but local cost realities should override the defaults.
Concrete action: Pull the underlying indicators, re-run the composite for your client country at higher granularity (sub-national or sectoral), and feed findings back to the partnership team.

Caveats & responsible use

  • Composite scores (0–100) are normalized indicators, not absolute rankings. A 65 means "strong relative to the 30-country sample," not "65% of some real-world maximum."
  • Cross-country comparisons should account for income group, population, and economic structure. A score of 20 in Ethiopia (pop. ~126M) is a different policy problem than the same score in a smaller economy.
  • The Simulator's projections are illustrative — cost-per-point and workers-per-$M are calibrated to global benchmarks. Local labour-market realities (currency, wages, existing subsidies) should override defaults before any commitment is made.
  • Gender data covers 25 of 30 countries — Rwanda, Haiti, Sudan, Guinea-Bissau, and Myanmar are not included in the GSMA Mobile Gender Gap Report 2024 sample. Treat gender-tab insights as directional for those countries.
  • The dashboard is a decision-support tool, not a replacement for country-level qualitative analysis or stakeholder consultation.

Key Citations

World Bank — Digital Jobs and Skills Brief
worldbank.org — Digital Jobs and Skills
Overview of World Bank digital skills programs and research.
World Bank — Gender and Inclusion in Digital Development
worldbank.org — Gender and Digital Inclusion
Gender digital divide data: participation rates, mobile internet access gaps.
Martins-Neto et al. — Click, Code, Earn: The Returns to Digital Skills (2025)
SSRN — Click, Code, Earn
Cross-country evidence on wage returns to digital skills using 67M+ job postings from 29 countries (2021–2024).
World Bank — Digital Progress and Trends Report 2025: Strengthening AI Foundations (2025)
openknowledge.worldbank.org — full PDF  ·  stable handle  ·  landing page
Primary source for the Market Demand tab. Lightcast online-vacancy analysis (Chapter 5: Competency — Digital Skills, pp. 71–93) including the 1.7M global AI vacancies figure, 9× GenAI surge, 25–36% wage premium for GenAI literacy, sectoral demand shares, and basic-digital-skills supply gaps by income group. Publisher: International Bank for Reconstruction and Development / The World Bank, Washington, DC.
World Bank Data360 — Digital Skills Indicators
data360.worldbank.org
10,000+ indicators including education, digital literacy, workforce data.
Stanford HAI — AI Index 2025
hai.stanford.edu — AI Index 2025
Business AI adoption rates (78%, up from 55% in 2023), workforce impact, global skills trends.
World Bank — World Development Indicators: Education
databank.worldbank.org — WDI
Tertiary enrollment, STEM graduates, education spending by country.
UNCTAD — Digital Economy Reports
unctad.org — Digital Economy
Digital skills components of eTrade assessments for 36+ countries.
Oxford Insights — Government AI Readiness Index 2025
oxfordinsights.com — GARI 2025
Skills as one of three key dividing lines in government AI readiness.
World Bank — What Works to Advance Women's Digital Literacy? A Review of Good Practices and Programs (2025, P173166)
documents.worldbank.org — full PDF
Source for the named real-world exemplars surfaced in the Intervention Simulator: Intel She Will Connect (Box 6, p.46), BRAC Shakti (Box 17, p.58), GSMA MISTT / Baxnaano Somalia (Box 27, p.82), Community Agent Network Philippines (Box 32, p.88). Includes the global mapping of digital-literacy programmes (Appendix A) and the multidimensional digital-literacy framework (Appendix B).
World Bank — Africa Centers of Excellence (ACE I, II, Impact) 10-year review (2025)
ace.aau.org — Africa Centers of Excellence  ·  worldbank.org — ACE programme page
Source for the University STEM Programs anchor in the Intervention Simulator: $729M total financing ($657M IDA + $72M AFD), 90,000+ students enrolled (7,650 PhD · 30,200 Masters · 52,000 short professional), 80+ centers across 50+ universities in 20 African countries, 32% female enrollment.
Andela & CNCF–Linux Foundation Education — Africa Developer Training reports (2024)
andela.com  ·  linuxfoundation.org — Kubernetes Africa announcement
Source for the Coding Bootcamps anchor: ~110,000 African technologists trained via the Andela Learning Community (2014–2024), original fellowship client billing $50K–$120K / developer / year, current free Kubernetes track of 5,600+ in cohort 1.
Moringa School Kenya — public outcomes reports
moringaschool.com
Source for the Coding Bootcamps anchor: 8,000+ alumni, 7,000+ trained, 85% job-placement rate within 6 months, tuition ~$1,200–$1,560 USD per learner; TVETA-accredited.

Built by Georgetown University students · Data Science & Public Policy Practicum · Spring 2026
In collaboration with the World Bank Digital Development Team · Delivery target: Early–Mid May 2026