WHAT INDIA SPENDS
ON PUBLIC LIBRARIES.
Kulkarni, Balaji & Dhanamjaya (2025) put the number at ₹11.62 per Indian per year (nominal) for 2018-19 — the highest point on record. We re-ran their analysis with the government's own annual population projections (TG 2020) and the MoSPI implicit GDP deflator (NAS, base 2011-12). After those two corrections, and after extrapolating state-level CAG accounts through 2024-25, the real story is this:
RISE, PEAK, RETREAT.
Between 2014-15 and 2018-19, state expenditure on public libraries (budget head MH 2205-105) nearly doubled in nominal terms. Adjusted with the MoSPI implicit GDP deflator (base 2011-12), the real-terms gain was modest but real: from ₹4.16 to ₹7.87 per Indian per year. From 2019-20 onwards, the line bends down — and according to the trend in the underlying CAG accounts, it has not bent back up. By 2024-25, almost all the post-2014 real-terms gain has retreated.
NATIONAL PER CAPITA EXPENDITURE ON PUBLIC LIBRARIES · ₹/PERSON/YEAR · INCLUDES UNION GOVT. SOLID = CAG ACTUALS (2014-21). DASHED = CAGR EXTRAPOLATION FROM 2021-22 (SEE METHODS). THE 2019-20 DROP IS DRIVEN BY WEST BENGAL CESS RECLASSIFICATION, NOT POLICY.
FROM ₹0.10
TO ₹87.94.
31 states and UTs across nearly four orders of magnitude. Jharkhand at one end. Goa at the other. The bottom half clusters near zero; the top is stretched thin. Read the strip below before the regional breakdown — and the two phenomena that distort everything that follows.
FIVE TRAJECTORIES.
Five states wrote library cess into their Public Libraries Acts — a small statutory levy, ring-fenced for libraries outside the annual budget. Tamil Nadu in 1948, before the Constitution. Kerala in 1989, just before liberalisation closed the door. Tamil Nadu's line is steady. The other four are not. West Bengal's expenditure jumps 35× between adjacent years. The cess does not stabilise library funding. It makes the line unreadable.
OFF THE SCALE.
Goa has 1.6 million people — about one outer Mumbai suburb — tourist revenues that outpace its population, and a Portuguese inheritance in which libraries are civic ornament. The press cites the unweighted state average of ₹11.33 as "India's library spend." Goa alone is what pulls that number above the population-weighted reality of ₹7.87 (2018-19 real, 2011-12 ₹).
THE STATE'S OWN MAP.
In 1956, Nehru proposed grouping the reorganised states into Zonal Councils — advisory bodies meant to "develop the habit of cooperative working" between linguistically and culturally adjacent States. The States Reorganisation Act, 1956 set up five such Councils. The North Eastern Council was added by statute in 1972. These are the State's own regional categories. Read the library line against them and the picture sharpens — particularly in the Central zone, where the State has grouped its largest population block with its lowest library investment.
- Uttar Pradesh₹0.70
- Madhya Pradesh₹0.81
- Chhattisgarh₹0.93
- Uttarakhand₹1.33
- Punjab₹0.73
- Haryana₹1.10
- Rajasthan₹1.10
- Delhi₹3.19
- Himachal Pradesh₹8.63
- Jammu & Kashmir₹9.74
- Jharkhand₹0.10
- Bihar₹0.27
- Odisha₹0.90
- West Bengal₹14.58
- Gujarat₹3.72
- Maharashtra₹9.20
- Goa₹87.94
- Telangana₹10.34
- Tamil Nadu₹12.02
- Kerala₹18.26
- Andhra Pradesh₹18.32
- Karnataka₹26.08
- Puducherry (UT)₹37.47
- Nagaland₹2.47
- Assam₹3.49
- Manipur₹5.09
- Tripura₹7.85
- Meghalaya₹8.23
- Mizoram₹12.44
- Sikkim₹14.57
- Arunachal Pradesh₹35.56
The MHA's own classification, read against its own budget data, produces a verdict the State does not advertise: the zone the Centre treats as most fiscally dependent (NEC) spends the most per capita on public libraries; the zone holding over a quarter of India (Central) spends the least. Read together with Section 5's legal-architecture breakdown — every Central-zone state is in the no-Act group — this is not coincidence. It is the residue of seventy-six years of legislative choice.
TWO MAPS.
ONE BUDGET LINE.
The State of India has drawn its regional map twice. The Ministry of Home Affairs in 1956 grouped reorganised states into five non-overlapping Zonal Councils for administrative cooperation. The Ministry of Culture in 1985 drew seven overlapping Cultural Zones for "preserving and promoting" culture. Neither classification is neutral — each is a claim about what counts as a region of India. The charts below show one bubble per zone in each scheme: position is the zone's average ₹/person, area is the zone's population share. The leftmost red bubble — the underspend zone — stays red across both classifications.
EACH BUBBLE = ONE ZONE · X-POSITION = ZONE'S AVG ₹/PERSON (LOG SCALE) · BUBBLE AREA = ZONE'S SHARE OF INDIA'S POPULATION · COLOUR = TRAFFIC-LIGHT BY AVG ₹ (RED = LOWEST, GREEN = HIGHEST) · HATCHED ON RIGHT = MoC ZONES OVERLAP, SO MEMBER SHARES SUM > 100% · HOVER FOR ZONE DETAIL.
On the left chart, the deep-red bubbles belong to the MHA's Central Zonal Council: UP, MP, Chhattisgarh, Uttarakhand — four states, ~361 million people, 27% of India, all under ₹2 per capita. On the right chart, those same four bubbles are still red, but they're now joined by Bihar, Delhi, Haryana, and Rajasthan: the MoC's North Central Cultural Zone, headquartered at Prayagraj (Allahabad), covering ~584 million people, 41% of India. The classifications differ; the underspend doesn't.
The MoC's North Central Cultural Zone — headquartered at Prayagraj (Allahabad) — captures the underspend even more widely than MHA Central. It pulls in Bihar (which MHA places in the East), Delhi (MHA: North), Haryana (MHA: North) and Rajasthan (MHA: North), on top of UP / MP / Uttarakhand. Result: ~584 million people — 41% of India, headquartered at the Hindi belt's cultural capital — averaging ₹1.21 per person per year on public libraries.
And the MoC's classification reveals a second pattern the MHA grouping hides: several states sit in multiple cultural zones. Rajasthan is in three (North, North Central, West). Andhra Pradesh, Karnataka, Telangana straddle South and South Central. Assam, Manipur, Tripura sit in both East and North-East. The State can't settle on which "culture" each state belongs to — but settles, in both maps, on the same underspend.
STATES SITTING IN MULTIPLE MoC CULTURAL ZONES · RED ROWS = STATES IN 3+ ZONES.
FOUR STATES.
FOUR ANSWERS.
CAG accounts for 2021-22 onwards are not yet published for most states. For Assam, Goa, Rajasthan, and Odisha we extracted figures from state demand-for-grants documents through 2024-25. These four are not a representative sample. They are what we have. Read them as four answers to one question: when CAG actuals catch up, will the post-2018 retreat bend upward, stay flat, or keep falling?
Three of these four states' answers to the question are: flat or token. One — Goa — kept climbing but does so under conditions that do not generalise. The fourth — Assam — might break upward, but only if a Budget Estimate becomes an actual disbursement, and only by ~22% above what CAG would report. This is the empirical anchor for the national extrapolation in Section 1. It is consistent with retreat, not recovery.
ACT. CESS. TAX.
ABSENCE.
Spending is downstream of legislation. Before a State can fund public libraries sustainably, it has to commit, in statute, that public libraries are a thing the State maintains. India has had three generations of state library legislation — and 76 years into the Republic, a fourth group: 16 states, ~700 million people, no library law at all.
Read the numbers in Sections 1–3 against this legal architecture and the data speaks differently. Acts don't guarantee spending. Cess doesn't guarantee stability. The "free libraries" Act doesn't guarantee free libraries. But the inverse holds reliably: the states that never legislated public libraries are the states that don't fund them.
Five states. All wrote a library cess into their Public Libraries Act: a small statutory levy — typically on property tax or stamp duty — collected by local bodies and dedicated, by law, to library development. Tamil Nadu did this before the Constitution: the Madras Public Libraries Act 1948 was drafted with Ranganathan's input. Kerala did it in 1989 — two years before liberalisation closed the door on this kind of legislation.
These are the only states in India that have, in law, treated library funding as a public obligation removed from annual budget politics. They also produced the most volatile expenditure data (see Cess Paradox, Section 2) — because the visibility of cess in CAG books is itself opaque. But the commitment is there. It is reversible only by legislative action, not by a budget cut.
Four states. Wrote Acts but chose tax-funding — library spending must compete with everything else in the annual budget, every year. Maharashtra's 1967 Act built on the Kolhapur Princely State's 1945 Public Libraries Act (the earliest on the subcontinent). Telangana inherited the AP 1960 framework on bifurcation and enacted its own version in 2015.
Haryana 1989 is the anomaly. Section 2(e) of the Haryana Public Libraries Act defines a public library as one that "permits members of the public to use it for reference or borrowing without charging fee or subscription." This is the only Indian Act that legally defines public libraries as free. It passed two years before liberalisation. Haryana's per-capita library spend: ₹1.10.
Six states. Every Library Act passed in India after 1990 has been tax-funded. No state since Kerala 1989 has written cess into a public libraries Act. Strip the two small-population outliers (Goa, Arunachal) and the generation-III average is ₹2.32 per person per year — barely distinguishable from the no-Act group.
The Chattopadhyay Committee (1986) had already recommended a National Public Libraries Act with dedicated funding. Then 1991 happened. The National Knowledge Commission (2007) reframed libraries as "knowledge economy infrastructure" — digital, technocratic, not a constitutional public good. NEP 2020 reduced libraries to "one nation, one digital library." A Generation-III Library Act in this context is paper. Acts without funding-flow architecture.
Sixteen states. Of the eight lowest-spending in India, six are in this group: Jharkhand ₹0.10, Bihar ₹0.27, UP ₹0.70, Punjab ₹0.73, MP ₹0.81, Rajasthan ₹1.10. UP alone — 230 million people, more than the population of Brazil — has no statute naming public libraries as something the State maintains.
The data refuses a clean story even here. Arunachal Pradesh, which the live government roster has been slow to update with its 2009 Act, spends ₹48 per capita. Mizoram, no Act, spends ₹12.44. Some no-Act states (the NE small ones) spend well; the large no-Act states all spend nothing. The absence is not a uniform condition — but it is a reliable predictor of underspend at scale.
76 years. Tamil Nadu legislated in 1948. The Constitution commenced in 1950. UP has had three-quarters of a century to write a Public Libraries Act and has not. Bihar — the state of Nalanda — has not. Madhya Pradesh has not. Punjab — where the Phulkian princes founded state libraries in the 1880s — has not. This is not an oversight. It is, after three generations of opportunity, a positive political choice.
Read this way, the per-capita data in earlier sections becomes a map of the State's legislative choices over four generations. The cess states wrote the strongest commitment and produced the most opaque accounting. The post-1991 Act states wrote soft commitments and produced flat budgets. The no-Act states wrote nothing and spend the least. There is one direction available to the Indian State now that it has not tried: a National Public Library Law, drafted under Chattopadhyay (1986) and again in 2024, never enacted. Whether such a law would compel funding, or simply create a federal version of Haryana's "free" Act, is the open question.
A DECADE.
NO REAL PROGRESS.
Public libraries are not a luxury. They are one of the few genuinely universal public goods — a place where a child from a poor family reads the same book as a child from a rich one. India's public library system, where it exists at all in the large northern states, is chronically underfunded.
The ₹11–12 figure that has circulated as a benchmark came from Balaji et al.'s careful 2025 study. Corrected for a growing population and rising prices using the MoSPI implicit GDP deflator (base 2011-12), India's peak real expenditure was ₹7.87 per person in 2018-19. By 2024-25 that figure had retreated to ₹4.66 — 41% below peak, almost entirely undoing the post-2014 real-terms gain.
For scale: the central government alone spends over ₹1,000 per capita per year on defence. Combined Union and State spending on education and culture runs to ₹9.2 lakh crore — over ₹650 per Indian (Economic Survey 2024-25). The entire Ministry of Culture — museums, monuments, archaeology, performing arts — received ₹3,261 crore in 2024-25: ₹23 per Indian. The one national body tasked with supporting public libraries across all states received ₹27 crore in its last verified Parliamentary allocation: under one rupee per Indian. ₹4.66 for libraries is not a rounding error in that arithmetic. It is what a State that has decided libraries do not matter looks like.
The data also shows that this is not inevitable. Kerala, Tamil Nadu, Andhra Pradesh, and Karnataka built genuine library infrastructure over decades. Several small north-eastern states punch well above their weight. The political will exists somewhere in India. It just does not extend to the states where most Indians live.
Pages 1–16 are the analysis. Pages 17–18 are sources, methodology, and the per-state CAGR table — choose "pages 1-16" in your print dialog to skip the methodology appendix.
Sources, data & citations
What we used and where it came from. State-level expenditure data for 2014-15 → 2020-21 from Kulkarni, Balaji & Dhanamjaya (Source 1 below). State-level expenditure data for 2021-22 → 2024-25 from state demand-for-grants documents (Assam, Goa, Rajasthan, Odisha) accessed via Open Budgets India (Source 3). Annual state-level population from the MoHFW Technical Group's projections (Source 2). All replication checks against Kulkarni-Balaji-Dhanamjaya's 2018-19 per-capita figures pass within ±₹0.20 of their published nominal values.
Deflation to 2011-12 ₹. All real-terms per-capita figures on this page are deflated using the MoSPI implicit GDP deflator derived from National Accounts Statistics (GDP at current prices ÷ GDP at constant 2011-12 prices), using the latest published revision per year — Third Revised Estimates 2014-15 to 2020-21, Second Revised 2021-22, Final Estimates 2022-23, First Revised 2023-24, Provisional Estimates 2024-25. This is the standard deflator for real government-spending analysis used in the IMF Government Finance Statistics Manual (GFSM 2014), OECD Government at a Glance, World Bank Public Expenditure Review methodology, and the RBI's annual State Finances: A Study of Budgets. Single-year state and zone cross-sections are deflated by the 2018-19 factor (0.7404, i.e. a cumulative GDP-deflator level of 1.3506 since the 2011-12 base). Per-state inflation differs slightly from the national factor; we use the national one for cross-state comparability. Kulkarni-Balaji-Dhanamjaya's original published nominal figures are recoverable by dividing by the same factor.
Sources
- Kulkarni, S. R., B. P. Balaji, and M. Dhanamjaya. "What Is the Per Capita Expenditure on Public Libraries in India? An Empirical Analysis." In Book of Abstracts: Global Library Summit on Library Diplomacy, South Asian University, New Delhi, 5–7 February 2025, 125–32. New Delhi: LIS Academy, 2025.
- National Commission on Population. Population Projections for India and States, 2011–2036: Report of the Technical Group on Population Projections. Chaired by K. S. James. New Delhi: Ministry of Health and Family Welfare, Government of India, July 2020. nhm.gov.in/New_Updates_2018/Report_Population_Projection_2019.pdf.
- CivicDataLab and Centre for Budget and Governance Accountability. "Open Budgets India." Accessed May 2026. openbudgetsindia.org. State demand-for-grants documents for Assam, Goa, Rajasthan, and Odisha (2021–22 through 2024–25) accessed through the OBI corpus.
- Ministry of Statistics and Programme Implementation, Government of India. National Accounts Statistics: Gross Domestic Product at Current and Constant 2011-12 Prices. Annual series, base year 2011-12. Implicit GDP deflator computed as GDP-current-prices ÷ GDP-constant-2011-12-prices. Latest published revision per year used (Third Revised Estimates 2014-15 to 2020-21, Second Revised 2021-22, Final Estimates 2022-23, First Revised 2023-24, Provisional Estimates 2024-25). Accessed via the MoSPI public API, May 2026.
- Reserve Bank of India. State Finances: A Study of Budgets of 2025-26. Mumbai: RBI, January 2026. Statement 31: Social Sector Expenditure — All States and UTs combined: ₹26,54,955 crore (₹26.5 lakh crore) in 2023-24.
- Ministry of Finance, Government of India. Economic Survey 2024-25, Chapter 11 (Social Sector). Combined Union and State education and culture expenditure FY25 (BE): ₹9.2 lakh crore. Ministry of Culture total allocation 2024-25: ₹3,260.93 crore (SC 376th Report, March 2025; PIB PRID 1895562). RRRLF allocation FY 2021-22: ₹27.07 crore (SC 310th Report, February 2022, Table of Central library institutions).
Caveats. Budget Estimates and Revised Estimates are not actuals; actual spending may differ. Assam state-budget figures likely undercount by ~22% versus the CAG methodology. For states without published 2021–25 figures we use CAGR (2016-17 → 2020-21) extrapolation from the 2020-21 base — see the assumptions block below.
Extrapolation assumptions for replication
These figures are estimates. They indicate trajectory, not confirmed actuals. Anyone replicating should verify against state budgets as they become available.
Method: For each state, compute compound annual growth rate (CAGR) of CAG library expenditure over 2016-17 to 2020-21 (4-year window). Project forward: Exp(t) = Exp(2020-21) × (1 + CAGR)(t − 2020-21).
Anomalous states (West Bengal, Karnataka, Telangana): included in national totals but flagged red below; cess reclassification and formation-year effects make their CAGRs unreliable for individual trend claims.
| State / Body | CAGR 16-21 | 2020-21 ₹L base | Notes |
|---|---|---|---|
| Andhra Pradesh | +8.9% | 10,486 | — |
| Arunachal Pradesh | +4.7% | 901 | — |
| Assam | +8.4% | 1,711 | Actuals 2021-24 override |
| Bihar | +3.0% | 262 | — |
| Chhattisgarh | +7.5% | 312 | — |
| Delhi | −2.5% | 579 | — |
| Goa | +8.8% | 2,050 | Actuals 2022-25 override |
| Gujarat | +10.7% | 3,276 | — |
| Haryana | +36.3% | 1,136 | 2020-21 spike inflates CAGR |
| Himachal Pradesh | +9.9% | 539 | — |
| Jammu & Kashmir | −1.7% | 1,245 | — |
| Jharkhand | −27.3% | 33 | Trends to zero |
| Karnataka ⚠ | −10.0% | 7,321 | Cess spike 2018-19 |
| Kerala | −14.5% | 2,591 | Cess volatility |
| Madhya Pradesh | +13.8% | 1,066 | — |
| Maharashtra | −2.6% | 11,399 | — |
| Manipur | −2.5% | 171 | — |
| Meghalaya | +9.7% | 377 | — |
| Mizoram | +6.1% | 201 | — |
| Nagaland | +5.8% | 92 | — |
| Odisha | +3.8% | 497 | Actuals 2022-25 override |
| Puducherry | −2.1% | 707 | — |
| Punjab | −3.2% | 254 | — |
| Rajasthan | +4.2% | 1,098 | Actuals 2023-25 override |
| Sikkim | +7.2% | 142 | — |
| Tamil Nadu | +7.6% | 13,094 | — |
| Telangana ⚠ | +6.9% | 5,278 | Formation-year effect |
| Tripura | −70.6% | 3 | 2020-21 recording error |
| Uttar Pradesh | +32.9% | 2,314 | — |
| Uttarakhand | +2.6% | 194 | — |
| West Bengal ⚠ | −4.0% | 14,855 | Cess reclassification |
| Union Government | −6.7% | 17,225 | Central library grants declining |