Two decades of governance examined through data.
Eras compared. UPA = May 2004–May 2014; NDA = May 2014–2026. Years on the May boundary (e.g. 2014) are transitional and shaded accordingly.
Data vintage. Figures reflect the latest official releases available as of mid-2026, drawn from the primary sources listed in each chapter’s “View Sources” drawer. Several series are periodically revised (e.g. the GDP base-year change to 2022-23); revised numbers can differ from those current at the time of an event.
Attribution principle. Where a programme spans both eras, credit is shared: foundations laid in one era and scaled in another are noted as such (Aadhaar/NPCI under UPA → UPI scale under NDA; renewable targets, road and GST programmes that cross the boundary). End-of-era snapshots also reflect policy lag — outcomes such as mortality, learning levels, conviction rates and electrification trail the policies that drive them, so a data endpoint is not wholly attributable to the government holding it.
Source scope. V2.0 additions use only official statistics — Government of India ministries, RBI, MOSPI, NCRB, Union Budget documents, World Bank, IMF, UN and WHO. Two qualitative radar axes (“Press/Speech”, “Institutions”) and some legacy items rely on external indexes outside this scope; these are flagged where they appear.
⚠ Disputed / limited-comparability data. Some series carry known caveats — COVID excess deaths vs the official toll, the NCRB 2017 category change, Swachh Bharat usage vs construction, Ujjwala refill rates, and the PLFS survey redesign (2017-18 onward). These are marked with a ⚠ note beside the relevant chart. No winner is declared; this remains a two-government story.
The Periodic Labour Force Survey began in 2017-18 (NDA era), replacing the earlier quinquennial NSSO rounds. Because definitions and frequency changed, PLFS figures are not directly comparable with pre-2017 employment data, so the series is shown only for the period the official survey covers. From 2025 the survey moved to a January–December cycle. Source: MOSPI — PLFS.
UPA laid Aadhaar's foundation and created NPCI. NDA scaled it into the world's most advanced digital payments ecosystem.
Aadhaar, NPCI formation, National Broadband Mission all happened under UPA. NDA's digital achievements built on this infrastructure. This is a two-government story.
UPA favoured rights-based entitlements (MGNREGA, RTI, RTE). NDA favoured direct transfer infrastructure (DBT, Jan Dhan, Ujjwala). Both invested massively.
Swachh Bharat: RICE Institute found toilet usage 47–95%. Ujjwala: NSSO data shows many couldn't afford refills. MGNREGA: NDA initially called it "monument to failure" but increased budget post-COVID. DBT: Saved ₹3.5L Cr per govt, but Aadhaar-linked exclusion errors documented.
Health spending remains below WHO benchmarks across both eras. COVID's second wave exposed systemic cracks. Mortality metrics improved consistently.
Access expanded under both. Learning outcomes remain poor — a crisis neither era resolved. NEP 2020 is ambitious but implementation is uneven.
Defence exports grew 23×. Border roads nearly doubled. But Galwan exposed vulnerabilities that remain unresolved.
The defence-exports chart above is consistent with the Ministry of Defence’s published export totals. International datasets (e.g. SIPRI) measure arms transfers on a different basis and may not match the MoD series. Source: Ministry of Defence.
India's renewable expansion is globally significant. But coal dependence persists at 67%, and air pollution kills 2.1M annually.
The power-mix chart shows installed capacity. Because solar and wind are intermittent, renewables’ share of actual electricity generated is materially lower than their share of capacity — coal still supplies most generated units. Capacity growth should not be read as an equivalent share of supply. Source: Central Electricity Authority (CEA).
NCRB data shows violent crime declining. But NCRB dropped "communal violence" category in 2017, making pre/post comparisons unreliable.
NCRB discontinued the standalone “communal/religious riots” category in its 2017 reclassification, and recording practices (zero-FIR, new sub-heads) changed over the period. Crime counts across the 2017 boundary are therefore not strictly comparable, and rising registered cases can reflect improved reporting rather than rising incidence. Source: NCRB — Crime in India (methodology notes).
No winner declared. Both eras have distinct achievements and failures.
Normalization. Each of the 13 sectors is scored 0–10 by mapping its headline metric onto a common scale relative to its plausible range over 2004–2026 (min–max style). Scores are editorial summaries of the underlying official data, not measured quantities, and reflect end-of-era position rather than a single year. Capacity sectors (highways, airports, renewables) and outcome sectors (poverty, health) are scored on different logics, so cross-axis comparisons are indicative only.
Per-axis source map.
* Disputed dimensions. “Press/Speech” and “Institutions” (and “Communal”, marked Disputed in the grid below) draw on external perception/freedom indexes that fall outside this documentary’s official-source policy. They are retained for completeness but are qualitatively weighted and carry lower confidence than the eleven data-backed axes. Their inclusion does not change the “no winner declared” conclusion.
India's story is not one government's triumph over another. It is a civilisation accelerating. Rights-based entitlements were UPA's contribution. Infrastructure scale and digital revolution were NDA's. India's next chapter demands what two decades couldn't fully deliver — courts that move at the speed of commerce, a manufacturing base that matches its ambition, and air clean enough to breathe in its own cities. No single government has delivered yet.