Public policy basics

In short

Public policy is the stated and acted-upon course of government action (or deliberate inaction) to address a public problem — expressed through Acts, rules, schemes, budgets and administrative orders. It is a core paper in BA Political Science and Public Administration under UGC CBCS, and the backbone of UPSC/State PSC GS-II (governance, welfare schemes), as well as entry tests for think tanks and development consultancies. Employers who test it include NITI Aayog and ministry internship/consultant cadres, state policy units (e.g. Kerala's and Telangana's planning departments), research organisations like CPR, CSDS, IDFC Institute, PRS Legislative Research, Sattva, Dalberg, J-PAL/IFMR, and the CSR arms of large firms. The topic is examined both descriptively (define policy, explain the policy cycle, compare models) and analytically (critique a scheme, read a budget figure, design an evaluation).

Explained

What public policy is — definitions and nature

• Thomas Dye: "whatever governments choose to do or not to do" — the classic definition; its strength is that non-decision is also policy.
• Harold Lasswell: policy science is the study of "who gets what, when and how"; he first set out the decision-process stages that became the policy cycle.
• Robert Eyestone: "the relationship of a government unit to its environment"; Y. Dror: policy as a major guideline for action.
• Key features: goal-oriented; made by legitimate public authority; backed by legal sanction and public money; a *course* of action, not a single decision; both action and inaction.
• Policy vs allied terms: a goal is the desired end; a policy is the chosen course; a programme/scheme is the operational vehicle (e.g. PM-KISAN); a project is a time-bound unit; law is one instrument of policy; administration executes it.
• Types: (a) substantive (what government does — education policy) vs procedural (how it is done — RTI, procurement rules); (b) distributive (benefits spread widely — subsidies), redistributive (transfer from one group to another — progressive taxation, reservation), regulatory (restrains behaviour — SEBI, TRAI, pollution norms), constituent/self-regulatory (creates institutions — formation of NITI Aayog); (c) Theodore Lowi's typology is exactly this distributive–redistributive–regulatory–constituent scheme and is frequently asked.

The policy cycle — stages and how each works

1. Problem identification and agenda setting: a condition becomes a *problem*, then reaches the agenda. Kingdon's multiple-streams model: a policy window opens when the problem stream, policy (solutions) stream and politics stream converge — e.g. the 2011–12 anti-corruption agitation putting the Lokpal on the agenda.
2. Policy formulation: options are drafted — by ministries, NITI Aayog, expert committees (e.g. Kasturirangan Committee for NEP 2020), draft Bills, pre-legislative consultation (2014 policy requires 30 days' public comment on draft legislation), and Parliamentary Standing Committees.
3. Policy adoption/legitimation: approval by Cabinet, Parliament or the State legislature; delegated legislation (rules and notifications) laid before the House; assent of the President or Governor.
4. Implementation: the ministry, state department, district administration, panchayat and parastatals convert text into service delivery; this is where most Indian policy failure occurs ("implementation gap").
5. Monitoring and evaluation: output and outcome tracking — NITI Aayog's Development Monitoring and Evaluation Office (DMEO), CAG performance audits, social audits under the rural employment guarantee framework, Outcome Budget statements.
6. Policy feedback, succession or termination: continuation, modification, merger or closure — e.g. the Planning Commission was replaced by NITI Aayog on 1 January 2015; the Twelfth Five Year Plan ended in 2017 and plans were replaced by strategy/vision documents.
7. Caveat: the cycle is a heuristic, not a description — in practice stages overlap, loop back and run simultaneously.

Models of policy making

• Rational (comprehensive) model: define the problem, list all options, evaluate every consequence, pick the option maximising net social benefit. Limitation: needs unlimited information, time and consensus on values.
• Bounded rationality / satisficing (Herbert Simon): cognitive and information limits mean decision-makers *satisfice* — pick the first option that is "good enough".
• Incrementalism (Charles Lindblom, "muddling through"): policy moves in small steps from the status quo through successive limited comparison; describes most budget making well, but is conservative and poor at radical change.
• Mixed scanning (Amitai Etzioni): broad scanning for fundamental choices plus incremental detail — a middle path.
• Garbage-can model (Cohen, March, Olsen): in "organised anarchies", problems, solutions and participants are dumped in and choices emerge from chance couplings.
• Group theory and elite theory: policy as equilibrium of interest-group pressure vs policy as preference of a governing elite handed down to masses.
• Institutional and systems models: Easton's input → conversion (political system) → output → feedback; institutional rules structure outcomes.
• Punctuated equilibrium (Baumgartner & Jones): long stability broken by bursts of rapid change — the 1991 liberalisation and the 2020 pandemic responses are standard Indian illustrations.
• Advocacy coalition framework (Sabatier): competing coalitions bound by core beliefs, with policy-oriented learning over a decade or more.

Policy instruments and the Indian institutional machinery

• Instruments (tools): legal/regulatory (Acts, standards, licences); economic (taxes, subsidies, cash transfers, cesses, tariffs, user charges); direct provision (schools, PDS, hospitals); information and persuasion (campaigns, labelling, advisories); institutional (creating a regulator or mission); and nudges (choice architecture — NITI Aayog's behavioural work; the Swachh Bharat campaign used norms and pledges).
• Executive: Cabinet and Cabinet Secretariat; Prime Minister's Office; ministries; Department of Expenditure for appraisal; the Public Investment Board/Cabinet Committee on Economic Affairs for large projects.
• Think tank: NITI Aayog (not a Constitutional or statutory body — created by Cabinet resolution) — Governing Council of all Chief Ministers, Vice-Chairperson, CEO; publishes the SDG India Index, Multidimensional Poverty Index, Aspirational Districts/Blocks rankings.
• Legislature: Parliament passes laws and the Budget; Standing Committees, Public Accounts Committee and Estimates Committee scrutinise; Question Hour and Private Member's Bills raise the agenda.
• Fiscal: Union Budget presented on 1 February; the Finance Commission (Fifteenth Finance Commission recommended devolution of 41% of the divisible pool to States for 2021–26; the Sixteenth Finance Commission's award period begins 2026–27) decides vertical and horizontal sharing; GST Council under Article 279A decides indirect-tax policy; the FRBM Act, 2003 frames fiscal targets.
• Judiciary: judicial review and public interest litigation shape policy — right to food, right to education, environmental rulings; courts increasingly also caution against judicial over-reach into policy choices.
• Regulators: RBI, SEBI, TRAI, CCI, CERC/SERCs, IRDAI, FSSAI — delegated rule-making is a growing share of policy.
• Federal and local: Seventh Schedule Union/State/Concurrent Lists determine who may act; the 73rd and 74th Amendments (1992) devolve to panchayats and municipalities (Eleventh and Twelfth Schedules); centrally sponsored schemes come with State cost-sharing.
• Non-state actors: civil society, media, business associations (CII, FICCI), unions, farmer organisations, think tanks and international bodies (World Bank, UN agencies).

Policy analysis and evaluation methods

• Policy analysis is prospective (ex ante — which option should we choose?); evaluation is largely retrospective (ex post — did it work?).
• Bardach's "eightfold path": define the problem; assemble evidence; construct alternatives; select criteria; project outcomes; confront trade-offs; decide; tell your story.
• Criteria: effectiveness, efficiency (cost per unit outcome), equity, administrative feasibility, political acceptability, legality, sustainability.
• Techniques: cost-benefit analysis (CBA), cost-effectiveness analysis (CEA), multi-criteria analysis, regulatory impact assessment, SWOT and stakeholder mapping, Theory of Change and the logframe (input → activity → output → outcome → impact).
• Evaluation types: *formative* (during, to improve) vs *summative* (after, to judge); *process* vs *impact*; social audit (participatory verification by beneficiaries, mandated in the rural employment guarantee law); concurrent evaluation.
• Causal identification: randomised controlled trials (J-PAL's Indian work), difference-in-differences, regression discontinuity, propensity-score matching — all attempt to construct a counterfactual.
• Data sources: NSSO/Periodic Labour Force Survey, NFHS, Census (the long-delayed next Census is reported to be scheduled for 2027), Economic Survey, RBI's *Handbook of Statistics on Indian Economy*, data.gov.in, PRS Legislative Research tracking.
• Transparency and accountability: Right to Information Act, 2005; CAG audits under Article 148–151; Citizen's Charters; the Digital Personal Data Protection Act, 2023 now constrains how beneficiary data is used.

Constraints, failure and reform in the Indian context

• Implementation gap causes: vague objectives, weak capacity and vacancies, fund-flow delays, poor last-mile coordination, elite capture, corruption, measurement by expenditure rather than outcome.
• Pressman and Wildavsky: the more clearance points a policy must pass, the lower the probability of faithful implementation. Top-down (Sabatier–Mazmanian) vs bottom-up (Lipsky's *street-level bureaucrats*, who exercise real discretion) perspectives.
• Policy failure types: design failure, implementation failure, political failure; unintended consequences (subsidised power encouraging groundwater depletion in Punjab and Haryana), moral hazard, targeting errors (exclusion vs inclusion error).
• Rationing and targeting: universal vs targeted provision; self-selection (manual work requirement in the employment guarantee); the Socio-Economic and Caste Census-based deprivation criteria.
• Reforms in practice: Direct Benefit Transfer with Aadhaar and Jan Dhan accounts; JAM trinity; GST (one national indirect tax; slabs rationalised to 5%, 18% and a 40% demerit rate from 22 September 2025); the four Labour Codes in force from 21 November 2025, replacing the Factories, Minimum Wages, Industrial Disputes and BOCW Acts; the Income-tax Act, 2025 replacing the 1961 Act from 1 April 2026; outcome budgeting; Aspirational Districts Programme; Mission Karmayogi for capacity building.
• Recent policy landmarks worth citing: National Education Policy 2020; Jal Jeevan Mission; Ayushman Bharat PM-JAY; PM Gati Shakti; National Logistics Policy 2022; the Women's Reservation Act (106th Amendment), 2023, whose implementation awaits a census and delimitation.
• Second Administrative Reforms Commission (2005–09) recommendations on ethics, e-governance and citizen-centric administration remain the standard reference for reform questions.

Key points

• Public policy — a purposive course of government action or inaction to address a public problem, backed by authority and public funds.
• Dye's definition — "whatever governments choose to do or not to do"; makes non-decision a policy.
• Lowi's typology — distributive, redistributive, regulatory, constituent policies, each with its own politics.
• Policy cycle — agenda setting, formulation, adoption, implementation, evaluation, feedback/termination; a heuristic, not a literal sequence.
• Kingdon's policy window — change when problem, policy and politics streams converge.
• Rational model — maximise net social benefit; defeated by information, time and value disagreement.
• Satisficing (Simon) — bounded rationality leads to "good enough" rather than optimal choices.
• Incrementalism (Lindblom) — small steps from the status quo; describes budgeting; weak on radical reform.
• Mixed scanning (Etzioni) — combine broad fundamental review with incremental detail.
• Easton's systems model — inputs (demands, supports) → conversion → outputs → feedback.
• Implementation gap — the difference between policy intent and delivered outcome; India's dominant policy problem.
• Street-level bureaucrats (Lipsky) — frontline staff whose discretion effectively makes policy.
• Policy instruments — regulation, taxes and subsidies, direct provision, information, institutions, nudges.
• NITI Aayog — non-statutory, non-Constitutional think tank replacing the Planning Commission on 1 January 2015; DMEO does evaluation.
• Fifteenth Finance Commission — 41% of the divisible pool to States for 2021–26; the Sixteenth's award starts 2026–27.
• Output vs outcome vs impact — toilets built (output), open-defecation-free status (outcome), fall in child stunting (impact).
• Exclusion vs inclusion error — eligible people left out vs ineligible people included; targeting always trades one against the other.
• RTI Act, 2005 and CAG audit — the two main instruments making policy accountable after the fact.

Formulas and examples

Example 1 — Cost-benefit appraisal of a rural road. A district road costs ₹8 crore to build in year 0 and ₹20 lakh a year to maintain for 10 years. Annual benefits (saved travel time, lower crop spoilage, better market access) are ₹1.6 crore a year for 10 years. Social discount rate r = 8%.

• Annuity factor for 10 years at 8% = [1 − (1.08)⁻¹⁰]/0.08. (1.08)¹⁰ ≈ 2.1589, so (1.08)⁻¹⁰ ≈ 0.4632 → factor = (1 − 0.4632)/0.08 ≈ 6.710.
• PV of benefits = 1.60 × 6.710 ≈ ₹10.74 crore.
• PV of costs = 8.00 + (0.20 × 6.710) = 8.00 + 1.342 ≈ ₹9.34 crore.
• NPV ≈ 10.74 − 9.34 ≈ +₹1.40 crore; BCR ≈ 10.74/9.34 ≈ 1.15 → accept, but the margin is thin, so test r = 12% before recommending.

Example 2 — Targeting errors in a food subsidy. A district has 2,00,000 eligible households. 1,70,000 of them hold ration cards, and audit finds a further 30,000 cardholders who are ineligible.

• Eligible but excluded = 2,00,000 − 1,70,000 = 30,000 → exclusion error = 15.0%.
• Total beneficiaries = 1,70,000 + 30,000 = 2,00,000 → inclusion error = 30,000/2,00,000 = 15.0%.
• Reading: the headcount matches the eligible population exactly, yet 30,000 deserving households get nothing — a coverage total can hide both errors, which is why error rates, not totals, are reported.

Example 3 — Cost-effectiveness of two literacy options. Option A: extra teachers, ₹2.4 crore for 4,000 children reaching grade-level reading. Option B: a remedial-volunteer programme, ₹90 lakh for 2,500 children reaching the same level.

• A: 2,40,00,000 / 4,000 = ₹6,000 per child. B: 90,00,000 / 2,500 = ₹3,600 per child.
• B is 40% cheaper per outcome. But A reaches 1,500 more children; if the budget is fixed at ₹2.4 crore, B would reach 2,40,00,000/3,600 ≈ 6,667 children — so B dominates, *provided* it scales and quality holds.

Example 4 — Difference-in-differences evaluation. Enrolment of girls in treatment blocks (cycle scheme) rose from 62% to 78%; in control blocks it rose from 60% to 68%.

• Treatment change = +16 pp; control change = +8 pp.
• Impact = 16 − 8 = +8 percentage points, the part not explained by the statewide trend. A simple before-after reading would have overstated the effect by a factor of two.

Example 5 — Devolution arithmetic. Divisible pool = ₹20,00,000 crore; States' vertical share 41%; a State's horizontal share from the Commission's formula = 6.0%.

• To all States = 0.41 × 20,00,000 = ₹8,20,000 crore.
• To that State = 0.06 × 8,20,000 = ₹49,200 crore (untied tax devolution, before grants and centrally sponsored scheme transfers).

Where it is used

• Civil services and government: policy desks in central ministries, State planning and finance departments, district administration; UPSC GS-II and Essay, State PSCs, and the Indian Economic/Statistical Services.
• Think tanks and legislative research: NITI Aayog consultants and Young Professionals, PRS Legislative Research, Centre for Policy Research, CSDS, ICRIER, Takshashila and IIM/ISPP policy programmes; LAMP Fellowship and Chief Minister's Fellowship schemes in several States.
• Development consulting and impact evaluation: Dalberg, Sattva, Samagra, MicroSave, J-PAL South Asia/IFMR, World Bank and UN country offices — roles in Theory of Change design, baseline surveys, RCTs and dashboards.
• Corporate public affairs and regulatory affairs: government-relations teams in telecom, pharma, energy and fintech firms tracking TRAI, CDSCO, CERC and RBI consultations; industry bodies CII, FICCI, NASSCOM.
• CSR and philanthropy: Section 135 CSR teams and foundations (Azim Premji Foundation, Tata Trusts, Piramal) designing and evaluating programmes aligned to government schemes.
• Journalism, data and advocacy: policy reporting, fact-checking budget and scheme claims, RTI-based accountability work, and NGO advocacy in health, education and environment.

Interview questions

• Q: Define public policy and explain why "inaction" counts as policy. A: It is a purposive course of government action on a public problem; a deliberate decision not to act — say, not regulating a sector — allocates costs and benefits just as a positive decision does, so it is a policy choice.
• Q: What is the difference between a policy, a scheme and a law? A: A policy is the chosen course and its objectives; a law is one instrument giving it legal force; a scheme is the operational programme with budget, guidelines and beneficiaries.
• Q: Walk me through the policy cycle. A: Agenda setting, formulation, adoption/legitimation, implementation, monitoring and evaluation, then feedback leading to continuation, change or termination — with the caveat that real policy loops and overlaps rather than running in order.
• Q: Why is the rational model rarely used in practice? A: It assumes complete information, unlimited analytical time and agreement on values; real decision-makers face bounded rationality, deadlines and contested objectives, so they satisfice or move incrementally.
• Q: Difference between output, outcome and impact? A: Output is what the programme produces (taps installed), outcome is the resulting change in behaviour or service use (households with piped water), impact is the long-run welfare change (fewer diarrhoeal deaths).
• Q: How does NITI Aayog differ from the Planning Commission? A: Both were created by Cabinet resolution, but the Planning Commission allocated plan funds through five-year plans, while NITI Aayog is an advisory think tank with no allocative power, working through State Chief Ministers and vision documents.
• Q: A scheme covers the right number of people but the poorest are missing. What is happening? A: Simultaneous exclusion and inclusion error — a correct headcount masks eligible households left out and ineligible ones included; report error rates and fix identification, not the total.
• Q: What is the implementation gap and name one way to narrow it? A: The shortfall between policy intent and delivered result, caused by vague goals, weak capacity and too many clearance points; narrowing it needs fewer decision points, frontline capacity and outcome-based monitoring rather than expenditure monitoring.
• Q: Why is a counterfactual essential to evaluation? A: Without an estimate of what would have happened anyway, before-after change is confounded by trends and other interventions — hence control groups, difference-in-differences or RCTs.
• Q: Which policy instrument would you choose to cut urban air pollution, and why? A: A mix: regulation (emission norms), economic tools (congestion or fuel pricing, scrappage incentives), direct provision (public transport) and information (real-time AQI), because a single instrument invites avoidance and has no revenue or behavioural complement.
• Q: How does federalism constrain policy design in India? A: The Seventh Schedule decides who may legislate; most welfare delivery is on the State or Concurrent List, so the Union acts through centrally sponsored schemes with cost-sharing, conditionalities and Finance Commission transfers.
• Q: Give an Indian example of an unintended consequence of a well-intended policy. A: Free or heavily subsidised farm power with assured procurement of paddy encouraged groundwater over-extraction in Punjab and Haryana, creating a water crisis the policy never targeted.
• Q: What role do courts play in Indian policy making? A: Through judicial review and PIL they have expanded rights (food, education, environment) and forced administrative action, though they also warn against substituting judicial preference for executive policy choice.

Common mistakes

• Treating the policy cycle as a literal, tidy sequence rather than an analytical heuristic, and so missing overlap, feedback loops and abandoned stages.
• Confusing policy with law, or naming a scheme when asked for the policy — PM-JAY is a scheme under health policy, not the policy itself.
• Calling NITI Aayog a Constitutional or statutory body, or saying it allocates funds to States.
• Reporting coverage totals or expenditure as evidence of success, with no outcome measure, counterfactual or error rate.
• Mixing up exclusion and inclusion error, or assuming reducing one does not raise the other.
• Attributing before-after change entirely to the intervention, ignoring secular trends and selection into the programme.
• Writing exam answers with only Western theorists and no Indian institution, Act, scheme or figure — or quoting stale ones (five-year plans as current, pre-2025 labour statutes, old GST slabs).
• Unit and arithmetic slips in appraisal questions — mixing lakh with crore, forgetting to discount future benefits, or quoting BCR without the discount rate used.

Related topics

• Political science careers · Indian Constitution: structure and federal features · Public administration: theories and organisation · Indian political economy and planning · Local self-government and the 73rd/74th Amendments · Governance, accountability and RTI.