The India Power Atlas
Part X · The Underlying Studies
Chapter 45

The Flow of Money: An Overview

The master flow dataset of the Atlas follows money from its source, through the bloc it strengthens, to its function and recipient. This chapter brings together the headline visuals built on it: the master and domain flow diagrams, the treemap of recipients, the time series by module and the networks of funders and recipients.

The 15 figures in this chapter are reproduced from the Atlas report and its synthesis, with the passage of the study that explains each. They keep the study’s own labels and colours: RW denotes the Hindutva-aligned bloc and LW the coalition that contests it (see A Note on Terms).

Figure 45.1. RW leads non-state money, mass exposure and Union institutions; the state moves the most money
RW leads non-state money, mass exposure and Union institutions; the state moves the most money

Source: synthesis/out (HEADLINE_flows, MC, T20); M5 discourse share; M8 institution census; S-A T56/T57.

Headline KPIs with confidence tag and classification basis (interactive version in the HTML report). Classification basis: Lens 1 (S0) unless stated.

Figure 45.2. Political, welfare and religious money dominate; most of it reaches GOV or unattributed recipients
Political, welfare and religious money dominate; most of it reaches GOV or unattributed recipients

Source: Atlas master flow dataset, headline year (synthesis/out/HEADLINE_flows.csv).

Source type → Bloc → Function → Recipient; raw ₹ crore. n = 5,216 flows. Classification basis: Lens 1 (S0, v1.1).

Figure 45.3. Political, welfare and religious money dominate; most of it reaches GOV or unattributed recipients (influence-adjusted)
Political, welfare and religious money dominate; most of it reaches GOV or unattributed recipients (influence-adjusted)

Source: Atlas master flow dataset, headline year (synthesis/out/HEADLINE_flows.csv).

Source type → Bloc → Function → Recipient; influence-adjusted ₹ crore-equivalent. n = 5,216 flows. Classification basis: Lens 1 (S0, v1.1).

Figure 45.4. Each domain draws on a few source types; bloc money concentrates in religion, politics and education
Each domain draws on a few source types; bloc money concentrates in religion, politics and education

Source: Atlas master flow dataset, headline year (synthesis/out/HEADLINE_flows.csv).

Source type → Domain → Bloc; raw ₹ crore. n = 5,216 flows. Classification basis: Lens 1 (S0, v1.1).

Figure 45.5. Each domain draws on a few source types; bloc money concentrates in religion, politics and education (influence-adjusted)
Each domain draws on a few source types; bloc money concentrates in religion, politics and education (influence-adjusted)

Source: Atlas master flow dataset, headline year (synthesis/out/HEADLINE_flows.csv).

Source type → Domain → Bloc; influence-adjusted ₹ crore-equivalent. n = 5,216 flows. Classification basis: Lens 1 (S0, v1.1).

Figure 45.6. GOV and unattributed money fill most of the map; LW and RW money is fragmented across many recipients
GOV and unattributed money fill most of the map; LW and RW money is fragmented across many recipients

Source: Atlas master flow dataset, headline year (synthesis/out/HEADLINE_flows.csv).

Bloc → domain → function → recipient; raw ₹ crore. n = 5,216 flows. Classification basis: Lens 1 (S0, v1.1).

Figure 45.7. Non-state money: LW is 24% of LW+RW raw and 34% after evidence-derived multipliers (Lens 1); state and market money lean further RW
Non-state money: LW is 24% of LW+RW raw and 34% after evidence-derived multipliers (Lens 1); state and market money lean further RW

Source: Atlas master flow dataset, headline year (synthesis/out/HEADLINE_flows.csv); MC_lw_rw_adjusted_quantiles.csv; synthesis/multipliers.md.

Panel A: share of all headline ₹ by bloc under three weightings. Panel B: LW share of LW+RW by sector, lens and weighting. n = 5,216 flows. Classification basis: Lens 1 (S0, v1.1) and Ensemble.

Figure 45.8. RW leads 8 of 11 layers (<45% LW); LW leads 1; near parity: L5 Legitimacy, L6 Identity
RW leads 8 of 11 layers (<45% LW); LW leads 1; near parity: L5 Legitimacy, L6 Identity

Source: synthesis/out/powerindex/T57_layer_balance.csv.

LW share of LW+RW power per hegemony layer L1–L11 (RW = 1 − LW); band = low–high range. n = 11 layers. Classification basis: Lens 1 (S0), expected value — S-A T57.

Figure 45.9. The 2015 jump is coverage, not change: religious-giving series (M2) begin; political money (M3) swung RW after 2014
The 2015 jump is coverage, not change: religious-giving series (M2) begin; political money (M3) swung RW after 2014

Source: synthesis/out/T4_module_fy_x_family_raw.csv.

Raw ₹ crore per fiscal year (FY end), module time series; breakpoints annotated. n = 9 modules. Classification basis: Lens 1 (S0, v1.1).

Figure 45.10. GOV money is service delivery and institutions; RW money is mostly religious upkeep, party and ads; LW money is spread wider
GOV money is service delivery and institutions; RW money is mostly religious upkeep, party and ads; LW money is spread wider

Source: Atlas master flow dataset, headline year (synthesis/out/HEADLINE_flows.csv).

Headline ₹ crore by function (cause) × bloc; shading is log scale; raw (left) vs adjusted (right). n = 5,216 flows. Classification basis: Lens 1 (S0, v1.1).

Figure 45.11. Religion, state promotion, knowledge and corporate money lean RW; education, ESG, law and media money lean LW
Religion, state promotion, knowledge and corporate money lean RW; education, ESG, law and media money lean LW

Source: Atlas master flow dataset, headline year (synthesis/out/HEADLINE_flows.csv).

LW share of LW+RW ₹ by domain, raw (●) vs influence-adjusted (○); right margin: LW+RW ₹ cr. n = 18 domains. Classification basis: Lens 1 (S0, v1.1).

Figure 45.12. RW-mapped parties ran 7 of these states in 2013 and 22 in 2026
RW-mapped parties ran 7 of these states in 2013 and 22 in 2026

Source: data/companion/M8_state_ruling_party_panel.csv; codebook/party_bloc_mapping.csv (T66).

Ruling party's bloc by state and year (T66 mapping); GOV = President's Rule. LW-mapped in 2026: Himachal Pradesh, Jammu & Kashmir, Jharkhand, Karnataka, Kerala, Telangana. n = 31 states/UTs × 23 years. Classification basis: T66 party mapping (Lens 2 per period).

Figure 45.13. Heatmap: state × layer balance
Heatmap: state × layer balance

Source: Atlas synthesis.

Figure 45.14. Money networks are hub-and-spoke: governments and aggregate private giving feed most large recipients
Money networks are hub-and-spoke: governments and aggregate private giving feed most large recipients

Source: Atlas master flow dataset, headline year (synthesis/out/HEADLINE_flows.csv).

Top 90 funder → recipient edges by headline ₹; node area ∝ ₹; edge width ∝ ₹; colour/pattern = recipient bloc. n = 126 nodes, 90 edges. Classification basis: Lens 1 (S0, v1.1).

Figure 45.15. Influence per rupee is set by function, not bloc: the diagonals are the multipliers
Influence per rupee is set by function, not bloc: the diagonals are the multipliers

Source: Atlas master flow dataset, headline year (synthesis/out/HEADLINE_flows.csv); S-A T55 durability.

Recipients: raw ₹ vs influence-adjusted ₹ (log–log); diagonals mark multipliers ×0.75–×7.5. n = 828 recipients ≥ ₹5 cr. Classification basis: Lens 1 (S0); user multipliers (U mid).