A global view. A country focus.
Explore policy signals, understand the scale of need, and look ahead to 2030.
19 October 2026
Explore 22 indicators covering age structure, dependency, reproductive health, education and employment. A larger working-age share creates potential; realizing a dividend also requires health, skills, decent work and inclusive institutions. Age dependency compares age groups, not actual economic dependency. Fertility and age structure are context variables, with no universal critical/high threshold. These indicators do not measure a realized economic dividend or estimate its monetary value.
Demographic structure has no population “affected” burden estimate. Each indicator retains its own denominator and source year. UNFPA Demographic Dividend Atlas ↗
Observed country values
Boundaries are illustrative and do not imply endorsement. Missing data are never treated as zero.
Where is the signal strongest?
Countries ranked within the current selection.
Read the evidence in context
One indicator record, organized by programme theme
Browse Child Poverty, Livelihood, Social Protection, PF4C, Cash-based Assistance, Population and Demographic dividend indicators, plus external economic and policy measures under Thematic. Shared indicator codes appear once; cross-theme relevance is shown on the same record. External measures open in their existing source module.
Country evidence matrix
Click a country for trends, source definitions and burden calculations.
Burden estimates
Method adapted from the CPD Intensity Dashboard: prefer published counts, then matched-denominator calculations, with planning proxies identified separately. Rate severity and population counts are independent.
Scenario year is used only in planning mode. Child-benefit basis applies only to child/family benefits. Household proxies concern households with children aged 0–17, a different unit and age definition from the ILO child calculation.
| Calculation |
|---|
Click ⓘ Method for formula, age/sex denominator, source, years, uncertainty and assumptions. Export CSV includes full calculation provenance. Counts across indicators overlap and must not be added.
Explore a country's path to 2030
Conditional ARIMA(1,1,0) with drift. At least 10 consecutive annual observations required; sparse series remain unforecasted.
Compare up to five countries
Choose countries yourself, use a peer classification, or explore neighbours of a reference country. One selection mode applies at a time.
Select countries · 0 / 5
How peer and neighbour suggestions work
World Bank peers share the income group or region in the dashboard’s country snapshot. UNICEF peers share a reporting region. OECD peers use the 2025 high or extreme fragility category; “not listed” is not treated as a resilience rating or peer group. These are classification peers, not statistically matched countries.
Neighbour suggestions prioritise land-border relationships from mledoze/countries (ODbL 1.0), then nearby alternatives ranked by great-circle distance using World Bank country coordinates. Peer suggestions use the same distance ordering, with alphabetical ordering when coordinates are unavailable. Missing coordinates are never treated as zero. Geographic proximity does not imply a shared border or similar policy context.
Five neighbours are suggested when available. The reference country counts toward the five-country comparison limit; deselect it to compare all five neighbours. Active country search, UNICEF region, World Bank region and income filters narrow every selection mode. Missing indicator values remain visible and do not affect which peers are suggested. Values retain their own reference years. Boundary relationships are descriptive and do not imply endorsement.
From global patterns to country action
Explore country-score distributions, regional patterns and income-group differences.
Medians describe scored countries, not the population of a region. Coverage differs by stage and group; no combined score across life stages is produced. Each stage uses its own disclosed basket and fixed global normalization.
Index settings & methodologyShow +Hide −
Life-cycle Development Index
Index reference year is independent of the single-indicator year filter. Country, income and region filters apply throughout this module. Normalization and ranks remain global. The source snapshot is fixed and is not changed by the single-indicator Refresh button.
Select four or five indicators
Explore the evidence behind each score
Country index matrix Show +Hide −
Click a country to open its index profile. Scores range from 0–100, higher is favourable. They are neither percentages of children developing well nor population counts.
| Weight sensitivity | Country profile |
|---|
Statistical diagnostics
Reproducible methodology
Design: four or five indicators per stage, selected from a disclosed life-stage basket. The design is exploratory and requires domain-expert review; no normalization is universally “best.” Basket membership, proxy ages and overlapping domains are visible. Adult and older-age composites are not produced because the dashboard lacks a validated basket of four comparable age-specific indicators for those stages.
Time alignment: choose each country's most recent observation within reference year − 4 through reference year. All selected components must be present and their years must span no more than three years. Otherwise the score is withheld. There is no imputation, extrapolation, hidden reweighting, or use of population projections.
Normalization: fixed global 2025 reference distributions use each country's latest observation in 2021–2025. A component requires at least 30 countries and a non-zero P5–P95 range. For favourable-high indicators: N = 100 × clip[(x − P5)/(P95 − P5), 0, 1]. For favourable-low indicators: N = 100 − the same transformed value. These analyst-chosen robust anchors limit outlier influence; P5/P95 are not policy targets. Values at the extremes are capped and ties remain ties. Baselines do not change when countries are filtered or the index year is changed.
Weighting and aggregation: index = ΣN/k, where k is four or five. Each component receives 25% or 20%. Equal weighting is a transparent normative choice, not an empirical proof of importance. Arithmetic aggregation allows strengths to compensate for weaknesses. No country population weighting is used. Global ranks use the mean rank for tied scores among complete cases; changing the basket or eligible sample can change ranks.
Robustness: compare the main score with midrank percentile normalization, geometric aggregation, leave-one-component-out results, and a ±20% change to one weight at a time followed by re-normalization. Report Spearman correlations and review pairs with |ρ| ≥ 0.85; correlated inputs can double-count a domain. These are diagnostic alternatives, not confidence intervals. Small differences or rankings should not be interpreted as statistically significant; source sampling errors and cross-country comparability are not fully propagated.
Audit: exports contain source values, years, normalized components, weights, contributions, reference bounds, missing-data reasons, sensitivity results, source retrieval dates, SHA-256 source hashes and the methodology version. Indicator definitions and original sources remain in the indicator library. Stage scores use different constructs and must not be directly compared as if they measured the same developmental outcome.
OECD / European Union / EC-JRC Handbook on Constructing Composite Indicators (2008) ↗ provides the methodological framework; it does not endorse this dashboard or its choices.
Official early childhood measurement · ECDI2030
ECDI2030 · SDG indicator 4.2.1
UNICEF's Early Childhood Development Index 2030 is the recommended measure in official UN SDG metadata for children aged 24–59 months. It uses 20 caregiver-reported developmental milestones covering health, learning and psychosocial well-being, with age-specific on-track thresholds. It is a survey-based measure and cannot be recreated from this dashboard's four/five country indicators.
Official ECDI2030 values are not loaded in this module. No substitute values or official-equivalence claims are generated. The former ECDI and ECDI2030 are not assumed comparable.
Official UN SDG 4.2.1 metadata ↗
UNICEF ECDI2030 instruments and technical resources ↗
Choose a visual that fits the evidence
Explore one variable, two paired variables, country comparisons, distributions, time trends, geography or valid age composition. Hover or focus a mark for its value. The same indicator retains the same unit across countries.
Chart indicator, year and alignment are independent of the single-indicator controls above. Country search, UNICEF region, World Bank region and income filters narrow the country list. Time charts show source history since 2000 through the chart reference year; the exact-year choice applies to cross-sectional, composition and bivariate charts.
Chart preview
Recommended charts and data-type rules
One variable
Bar, column, dot, lollipop and diverging bars compare numeric values. Histogram and box plot show distributions. Counts can use treemaps. Tables preserve exact values.
Two variables
Scatterplot uses matched-country, same-year observations. Bubble size adds a third variable: matching-year population. Different units stay on separate axes. No causal interpretation or regression is imposed.
Across time
Line, area, step line, small multiples, slope, dumbbell and country × year heatmap. Up to three countries share one indicator and common scale. Missing annual values stay visible as gaps.
Geography and composition
Classified choropleth and population-sized symbol maps. Pie, donut and 100% stacked bars are reserved for the complete, non-overlapping age structure: 0–14, 15–64 and 65+.
This is a dashboard-specific chart collection, not every visual available in Power BI. Sankey, network, Gantt, waterfall, funnel, radar and geographic drill-down are not offered because this dataset lacks the required flows, relationships, task dates, additive bridges, conversion stages, validated common-scale dimensions or subnational geometry. Adding those forms requires suitable data and definitions. Avoid 3D effects and dual-axis overlays that can obscure comparisons.
Official source API connections
Query configured indicators directly from World Bank, ILO, IMF and ADB. Live extracts retain observation periods, units, national sources and methodological notes. All existing World Bank indicators and six ILO social-protection functions have routes; provider variants are linked rather than silently replacing canonical series.
Select an indicator and countries, then request source data.
Only the dashboard owner can make source requests. Extracts are held in this browser session and do not overwrite published snapshots or population calculations. No login or API key is required for the tested public endpoints. A configured route does not guarantee data for every country or period.
Source history
One indicator, up to three countries. Missing and ambiguous periods are not interpolated. Hover or focus data points for exact values. IMF growth is derived using exact annual lags and a consistent reference base; no official IMF growth-series download is implied.
| Country | Period | Value | Unit | National source / status | Methodological notes |
|---|
Coverage and boundaries
World Bank: 108 annual catalogue indicators plus seven GEM series. ILO: six existing social-protection functions, quarterly unemployment and quarterly NEET. IMF: monthly CPI-based inflation and quarterly real-GDP growth. ADB: annual inflation and real-GDP-per-capita growth for covered Asia-Pacific economies. These are the configured ESP indicators, not every series in all four institutions’ databases.
The seven cohort and derived population series remain calculations or workbook imports; a live rate refresh cannot update their denominators. UNDP, Transparency International, IEP and OECD index/report editions retain their own import workflows. ILO source notes may include food benefits or households: no under-15 child denominator is assumed for these direct extracts. ADB annual observations are not relabelled quarterly. Older GEM API extracts cannot replace newer published snapshots.
Economic Monitor
High-frequency context for UNICEF economic and social policy. These are macroeconomic signals, not direct measures of child well-being.
Selected-country history
Compare up to three countries on one indicator and one scale. Hover or focus points for exact values. Gaps remain unconnected. The matrix below includes all countries matching this module’s region and search filters; missing data are never zero.
| Country | UNICEF region | Observed value | Period | Frequency |
|---|
Source selection and update policy
Connected: World Bank Global Economic Monitor, official monthly and quarterly observations. Its release schedule is separate from each country’s reporting frequency. The API provides snapshots to viewers and permits only the owner to attempt source refreshes. Successful refreshes update this browser session; published snapshots change only after a reviewed import and republication. There is no unattended refresh schedule.
Direct World Bank, ILO, IMF and ADB connections are available in . They preserve their own frequencies and source definitions. ILO quarterly labour indicators and IMF monthly CPI/quarterly GDP are available there; the selected ADB measures are annual. Requests do not imply automatic scheduled updates.
API: /api/economic?code=NYGDPMKTPSAKD&countries=KEN,IND,BRA. Add &refresh=1 for an owner-authorized source attempt. A failed or invalid response explicitly returns the dated snapshot.
Explore the global picture. Find the country evidence.
Colours indicate source-value ranges, not an official UNICEF classification. Latest available values may come from different periods. Boundaries are illustrative and do not imply endorsement; missing values are not zero.
Country index matrix
| Country | UNICEF region | Source value | Period | Frequency |
|---|
Definitions, source and interpretation
Explore selected-country history
Compare up to three countries on one indicator and scale. Hover or focus points for exact values. Gaps remain unconnected. Country history is independent of the map’s selected period.
OECD States of Fragility 2025
The primary country flag uses OECD’s published 18 extreme and 43 high fragility contexts. Other countries are labelled “Not on high/extreme list”; this does not establish whether a country was assessed or resilient. World Bank FY2027 institutional-fragility and FCV flags remain separate supplementary fields.
Report edition 2025 uses inputs from 2023 or earlier. Its tiers come from OECD’s multidimensional methodology, not the dashboard’s quartile bands. Dimension categories below are the published categories, not recalculated scores.
The complete OECD analytical input dataset is available from OECD on request. The 56-entry register below distinguishes existing or separately sourced series from unavailable OECD inputs. Newer source observations are not presented as the data used to derive the 2025 tiers.
OECD methodology and data availability ↗ · Indicator definitions ↗
61 high/extreme contexts · published dimension categories
| Country | OECD tier | Economic | Environmental | Human | Political | Security | Societal | WB institutional fragility | WB FCV |
|---|
Source: OECD’s linked States of Fragility interactive dataset, report edition 2025. Dimension categories: severe, high, moderate, low, minor. “Not listed” on a WB list is not a resilience assessment.
121 indicators. Seven connected themes.
Official series names, definitions and sources; availability varies by country.
From source data to decision signals
1. Source and coverage
World Bank World Development Indicators (source 2), with original custodians retained in each indicator's metadata. ILO social protection series are distributed through the UN SDG API where identified. Data cover 2000–2026 where released. Countries and economies follow World Bank classifications; aggregate regions are excluded. Latest values may refer to different years. Five recent observations are shown where available, not invented where absent.
2. Benchmarks and intensity
For most indicators, critical / high / moderate / low denote relative screening bands using the global distribution of latest available country values (quartiles), with the adverse direction specified in metadata. These are analytical thresholds, not official UNICEF classifications or universal standards. Context indicators are shown as distribution bands without an adverse interpretation. Ties share the same band.
Under-five mortality uses the SDG 3.2 target of 25 deaths per 1,000 live births; neonatal mortality uses 12. Low = at or below target, moderate = above target up to 1.5×, high = above 1.5× up to 2×, critical = above 2×. The target is official; the distance bands are analyst-defined. Global medians are unweighted country medians, not population-weighted global estimates.
3. Estimated population counts — CPD methodology
Published-first: imported UN IGME/UNICEF annual death counts and UNICEF–WHO–World Bank nutrition counts take precedence where available. WHO–UNICEF JMP counts use the coverage and population from the same source release. The observation year is preferred; otherwise the latest imported count retains its own year, explicitly separate from the displayed rate.
Analytical calculations use the correct age, sex, geography and reference year: prevalence ÷ 100 × eligible population; uncovered population = (100 − coverage) ÷ 100 × eligible population; event rates use their stated scale, for example adolescent births per 1,000 women aged 15–19. Under-five deaths are never inferred by multiplying under-five mortality probability by annual births.
Nutrition uses children aged 0–4; NEET uses youth aged 15–24, with sex-specific cohorts for female/male series. DTP3 uses the CPD age-1 planning proxy, not an official missed-vaccination count. ILO child-benefit estimates use ages 0–14 (birth to the 15th birthday), and remain planning estimates because source coverage may include household or food-benefit definitions. If a matching child denominator is unavailable, the household fallback is: population ÷ mean household size × share of households with a child aged 0–17 × coverage gap. Household inputs must come from the same source row at or before the population year; the source year gap and private-household-population approximation are disclosed.
The imported single-age population extract covers 141 CPD countries, 2000–2030. It came from the supplied population workbook; its revision and annual reference date were not identified. For existing all-age percentage calculations outside that extract, matching-year World Bank total population is the fallback. No equal-sex split or assumed school-age interval is used. World Bank gross completion ratios are not substituted for the CPD SDG attainment-cohort measure.
Published counts, analytical estimates, planning proxies, review-required calculations and constant-rate scenarios are labelled separately. Source-note qualifications can exclude calculations from source-compatible counts. All available estimates are visible by default; select Published / matched-source only for the narrower subtotal. Independent statistical review is not claimed. Missing values are not zero. Counts with different units are not summed or plotted on one scale; mixed-year subtotals are flagged. Never sum across indicators.
Source count uncertainty bounds are retained when supplied, without assuming a 95% level. No interval is invented for denominator or proxy uncertainty. Scenario mode changes the population year while holding the rate constant; it is not an ARIMA forecast. Imported published-count files and population datasets retain their CPD retrieval dates; refreshing an indicator rate does not refresh these source files.
4. Forecasting
The outlook fits ARIMA(1,1,0) with drift using conditional least squares on first differences, with |φ| capped at 0.98. It requires at least 10 consecutive annual observations, latest year 2020 or later. No missing years are interpolated. A three-year holdout is compared with a last-value forecast; a model that performs worse is withheld. The chart shows conditional 95% normal prediction intervals based on innovation variance; these exclude parameter and policy uncertainty. Percentage bounds are clipped to 0–100. Forecasts are exploratory, not causal predictions or official estimates.
5. Refresh and reproducibility
Only the authenticated dashboard owner can see or use Refresh. The server checks ownership before retrieving the selected series from its public API into this browser session. Derived population series are refreshed by rebuilding their source inputs, not by this button. The dated bundled snapshot remains the fallback if the endpoint is unavailable. Refresh does not silently overwrite the published dataset. CSV exports include country, year, series code, intensity and calculation status. Publication source includes the extraction scripts and full metadata.
6. Demographic dividend
The dedicated theme brings together 8 additional World Bank series and 14 existing indicators, without double-counting the shared indicators in the unique catalog. Working-age share and dependency ratios indicate age-structure context; fertility is not scored against a universal desirable level. Employment rates describe participation, not job quality. Family planning measures should be interpreted through voluntary access and reproductive rights. Context variables retain neutral distribution labels. See the UNFPA Demographic Dividend Atlas for the conceptual framework.
7. Population age groups and country classifications
The Population theme includes under 5 (0–4), 0–17, 10–19, 20–24, 6–10, 15–64, over 64 (65+), and under 15 (0–14). Cohorts overlap and must not be summed. Five age groups are sums of exact single ages in the CPD workbook; values are explicitly marked as estimates or projections. Ages 0–14 use the published WDI total. Ages 15–64 and 65+ are analytical counts from matching-country/year WDI age shares × total population, rounded to persons. Latest excludes future years; future CPD population values remain available in the year filter as source projections, never as observed data or fitted ARIMA forecasts.
Child Poverty includes deprivation and wider contextual indicators; all-population poverty measures retain their original universe and are not relabelled child-specific. PF4C means Public Finance for Children. Cash-based Assistance cross-lists cash-benefit measures from Social Protection; indicator definitions remain unchanged.
UNICEF reporting regions follow the geographic reporting list, not programme-office regions. Economies absent from that list remain unmapped. Primary fragility labels use the OECD States of Fragility 2025 high/extreme list. Absence from that list does not establish whether a country was assessed or resilient. Supplementary flags use two separate World Bank FY2027 lists effective 1 July 2026: institutional fragility and public FCV. “Not listed” is not a resilience rating or evidence of no conflict. These labels do not change indicator-intensity bands.
The observed-value map uses global classification thresholds, colours, adverse direction and tie rules. Numeric intervals in the legend show which values fall into each band. Global quartiles are computed for the chosen reference year (latest uses each country's latest available value), before country and region filters.
Official documentation
World Bank API documentation ↗
UN SDG indicator metadata ↗
ILOSTAT data ↗
SDG 3 targets ↗