Charts about the things we teach, built from public data.
Some we made ourselves from named datasets. Some are diagrams of the ideas we use at work. A few are famous old charts we rebuilt in code to see how they were made, and there's a shelf of newer work by other people worth your time.
Same numbers, two y-axes. Start a bar chart anywhere but zero and you can invent a trend or hide one. Every chart here starts at zero.
Principle: Tufte's "lie factor". A graphic should not say more, or less, than the data.
The Frameworks Wing
The ideas behind the numbers
Numbers tell you what happened. These are some of the models we use to ask why, and what to do next.
Arnstein's Ladder of Participation
Not all "participation" is equal. Sherry Arnstein's 1969 ladder climbs from token consultation to genuine citizen power — a staple of our governance and MEL teaching.
The Results Chain & the Attribution Gap
Every theory of change runs from inputs to impact. The hardest gap to cross — and to prove — is the one between what a programme produces and the change in people's lives.
The Poverty Trap
A reinforcing loop: low income starves savings and investment, which holds down productivity, which holds down income. Breaking any one link is the work of development.
Intersectionality
Disadvantage compounds. Gender, caste and class are not separate problems to add up — they intersect, and the overlap is where the deepest exclusion sits.
The Systems Iceberg
The events we react to sit above the waterline; the patterns, structures and mental models that produce them sit below. Lasting change works underwater.
Build your own
Every chart here was made with a browser and a public dataset. Pick a dataset, pick a frame, and try it yourself.
A chart is only as good as its source. Here is the dataset behind each one.
The great escapeWorld Bank Poverty & Inequality Platform (PIP), via Our World in Data — extreme poverty at $2.15/day (2017 PPP), 1990–2019.
India's childrenUN IGME / World Bank Open Data, indicator SH.DYN.MORT — under-five mortality per 1,000 live births, 1990–2022.
Poverty is not caste-blindNFHS-4 (2015–16) multidimensional poverty headcount by social group; national MPI ≈ 24.8%.
Whose carbon?EDGAR (European Commission / JRC) — territorial CO₂ emissions per capita by country, 2023.
The leaky pipelineUDISE+ 2021–22 (school) and AISHE 2021–22 (higher education) — female gross enrolment ratio by level.
How India keeps the lights onEmber / Central Electricity Authority — share of electricity generation, 2024.
The same gap, four waysNFHS-5, AISHE 2021–22, PLFS 2023–24, and Lok Sabha 2024 — by sex.
Two decades of progressNational Family Health Survey rounds 3, 4 and 5 — national indicators.
How hot, by 2100?IPCC Sixth Assessment Report (AR6, WG1, 2021) — projected warming by scenario, best estimate and very likely range.
India is heating upIndia Meteorological Department high-resolution gridded data (1901–2024), via Data for India — annual mean temperature anomaly.
Women re-enter the workforcePeriodic Labour Force Survey (PLFS), MoSPI — female LFPR, usual status, age 15+, 2017-18 and 2022-23.
Where the rupee comes fromUnion Budget 2025–26, "Budget at a Glance" (Ministry of Finance) — receipts and expenditure in paise per rupee.
Where India's poor actually liveNITI Aayog National Multidimensional Poverty Index, 2023 report (NFHS-5, 2019–21) — state headcount ratios, applied here to projected 2021 state populations.
One person, one vote?Census of India 2011 (state populations) with the current Lok Sabha seat allocation — seats frozen at the 1971 census by the 42nd Amendment, extended by the 84th.
India, one hexagon at a timeCensus of India 2011 — population by state, one hexagon per crore, rounded to the nearest crore.