Process and Techniques of Decision-Making: Models, Tools and the Narrowing of Choices
Simon called decision-making the heart of administration, and the syllabus makes it the opening entry of Administrative Behaviour. But the topic has two halves that most notes collapse into one. The process is a sequence — intelligence, design, choice, review — and a family of competing models running from rational-comprehensive through satisficing and incrementalism to the garbage can. The techniques are a distinct toolkit: cost-benefit and decision-tree analysis, operations research and game theory on the quantitative side; Delphi, nominal group and devil's advocacy on the qualitative. This guide covers both in depth, adds Allison's three lenses and the behavioural failures from groupthink to anchoring, decodes the 2023 stem, and maps every concept onto Paper 2 — Indian Administration.
Why This Topic Sits First in Administrative Behaviour
A decision is the selection of a course of action from among alternatives. Simon's proposition, opening Administrative Behavior (1947), is that decision-making is the heart of administration, and the vocabulary of administrative theory must be derived from the logic and psychology of human choice. Administration is not primarily structure, or rules, or morale — it is a continuous stream of choices, and everything else exists to shape them.
Public decision-making differs from private in three ways worth stating early. It is plural — as Seckler-Hudson observed, one individual may pronounce the decision but many contribute to reaching it, and it is part of the political system. It is exposed — subject to direct and indirect pressures a business decision-maker rarely faces. And it cannot be indefinitely deferred: a government cannot keep a flood, an epidemic or a court deadline pending while it gathers more information.
UPSC writes "Process and techniques of decision-making". Most notes cover only the first half — the sequence of steps and the competing models — and stop. The techniques are a separate body of content: cost-benefit analysis, decision trees, operations research, game theory, PERT and CPM, payoff matrices and Bayesian analysis on one side; Delphi, nominal group technique, brainstorming, devil's advocacy and dialectical inquiry on the other.
An answer that describes Simon and Lindblom and never names a single technique has answered half the question. Conversely, listing techniques without the models is mechanical. Prepare both, and signal the distinction explicitly in your structure.
Types of Decisions
The Process: Simon's Phases and the Operating Sequence
Simon's four phases remain the standard frame.
- Intelligence — searching the environment for conditions calling for a decision. The word is used in the military sense of surveying terrain, not the cognitive sense.
- Design — inventing, developing and analysing possible courses of action.
- Choice — selecting one course from those developed.
- Review, added later — assessing past choices so that outcomes feed the next cycle.
The question quoted the proposition that the process of decisions is largely a technique of narrowing choices. The word doing the work is narrowing, and the answer should show that decision-making is subtractive rather than additive.
How the narrowing happens, stage by stage. The universe of conceivable actions is first cut to the legally permissible — statute, rules and precedent eliminate most options before anyone deliberates. That set is cut again to the financially feasible by budget ceilings and delegation limits. It is cut again to the administratively implementable by available staff, time and capability. It is cut again to the politically acceptable. What reaches the deciding authority is rarely more than two or three surviving options — and by then most of the decision has already been made, invisibly, by whoever applied the earlier filters.
Whose theory this is. Simon's bounded rationality is the psychological version — limited sequential search stopping at the first satisfactory option. Lindblom's successive limited comparisons is the political version — only alternatives marginally different from the status quo are ever examined. Simon's decision premises is the organisational version — the organisation supplies the premises, and thereby narrows the field, before the individual reasons at all.
The critical turn to close on. If deciding is narrowing, then power lies with whoever controls the narrowing — the officer who drafts the Cabinet Note and decides which alternatives appear in it, the committee that shortlists, the guideline that fixes eligibility. This is Bachrach and Baratz's second face of power, and Lukes's third: the most consequential decisions are about what never reaches the table.
The Competing Models
The fastest way to make these models stop feeling abstract is to run one real Indian story through all of them. Take the evolution of the PDS.
The story in one line. A universal ration system became targeted in 1997, acquired computerised supply chains and digitised ration cards through the 2000s, was put on a statutory footing by the National Food Security Act 2013, was linked to Aadhaar for beneficiary authentication, and became portable across states through One Nation One Ration Card.
Rational-comprehensive would say: the objective was defined as food security, all delivery mechanisms were compared — universal PDS, targeted PDS, cash transfer, food coupons — and the optimum was selected. Nobody who has watched the process believes this.
Bounded rationality says: at each stage the decision-makers had partial information about leakage, no reliable count of the genuinely poor, and pressing time. They took the first design that looked good enough against the aspiration level of the moment — which is why targeting was adopted before the identification problem it depends on had been solved.
Incrementalism is the best single description of the whole sequence. Nobody redesigned PDS from first principles. Each step was a marginal adjustment to what already existed, comparable to the status quo and negotiable with the states — which is precisely why it survived changes of government that a comprehensive redesign would not have.
Mixed scanning reads the NFSA 2013 as the fundamental decision — a broad scan that reset the entitlement framework as a legal right — with everything after it, from card digitisation to portability rules, as incremental decisions inside that frame.
Dror would ask the uncomfortable question: did three decades of incremental adjustment ever confront the underlying design question of transfer in kind versus in cash? Incrementalism cannot pose that question, because the answer lies outside the range of alternatives adjacent to the status quo. This is exactly his charge that muddling through institutionalises inertia.
Garbage can notices that Aadhaar arrived as a solution before PDS was defined as its problem. A general-purpose identity platform, built for its own reasons, was available when the leakage problem came up for decision — and attached itself to it. Solutions in search of problems is not a cynical reading here; it is a fair description of how a platform finds its applications.
The Models in Brief
- Rational-comprehensive. The benchmark. Clarify values, list all alternatives, compute all consequences, select the optimum. It fails on every assumption, but it remains the standard against which the others define themselves — and it survives in cost-benefit appraisal, where a bounded version of it is genuinely used. Seen in India in the Detailed Project Report placed before the Public Investment Board, where alternatives, costs and benefits are formally compared before a large infrastructure investment is cleared.
- Bounded rationality and satisficing (Simon). Rationality is limited by knowledge, values, habit, time and computational capacity. The decision-maker searches sequentially and stops at the first alternative clearing an aspiration level that itself adapts — rising on success, falling on failure. Seen in India whenever a District Magistrate must decide before nightfall — relief camp locations after an embankment breach, chosen from buildings she already knows are sound and reachable, not from an optimal survey of the district.
- Incrementalism (Lindblom, 1959). The branch method against the root. Means and ends are chosen together; only marginally different alternatives are compared; the test of a good policy is that participants agree on it; and coordination comes from partisan mutual adjustment rather than central analysis. Seen in India in the annual budget, where last year's allocation is the starting point and the argument is about the margin; and in the decades-long reform of the PDS, MSP and fertiliser subsidy, each changed by increments negotiable with the states.
- Mixed scanning (Etzioni, 1967). Use the rational mode for fundamental, direction-setting decisions and the incremental mode for the detailed decisions within that frame — a weather satellite's broad-angle camera plus its zoom lens. Seen in India in the NEP 2020, which set the direction comprehensively while leaving curriculum, examination and institutional detail to be worked out incrementally by states and universities; and in GST, where the constitutional architecture was the fundamental decision and the Council's rate revisions are the bit decisions.
- Normative-optimum (Dror, 1968). Incrementalism is inherently conservative and suits only stable societies with satisfactory existing policies. Dror combines rational with extra-rational elements — judgement, intuition, creative invention — across eighteen phases in three stages: meta-policymaking, policymaking, post-policymaking. The most useful model for Indian answers, because it was written for societies that must transform rather than merely adjust. Seen in India in the 1991 reforms and in the design of India Stack — neither of which any amount of incremental adjustment to the existing system would have produced.
- Garbage can (Cohen, March and Olsen, 1972). In organised anarchies — unclear goals, unclear technology, fluid participation — problems, solutions, participants and choice opportunities float independently and couple largely by accident. Solutions search for problems as often as the reverse. Universities and, arguably, large committee-based governments qualify. Seen in India whenever a vendor's ready-made platform finds a problem to attach itself to during a budget cycle, or a proposal rejected by one committee resurfaces unchanged in another two years later because the participants and the timing have changed.
- Public choice. Buchanan, Tullock and Niskanen model officials as self-interested utility maximisers — the budget-maximising bureau — which makes decision-making an arena of rational self-interest rather than public purpose. Seen in India in the department that defends its budget line and its sanctioned strength irrespective of whether the function still needs them.
Allison's Three Lenses — The Same Decision, Read Three Ways
Graham Allison's Essence of Decision (1971), on the Cuban Missile Crisis, is the single most transferable framework in this topic because it shows that the model you use determines what you see.
| Model | The Unit of Analysis | The Question It Asks | Indian Application |
|---|---|---|---|
| I · Rational Actor | The government as a single unitary decision-maker with coherent goals | What objective made this the sensible thing to do? | The standard public explanation of any policy — "the government decided X to achieve Y" |
| II · Organisational Process | A constellation of loosely allied organisations, each with standard operating procedures | What organisational routines produced this output? | Why relief reaches a district in the form the existing SOP allows rather than the form the disaster requires |
| III · Bureaucratic Politics | Individual players bargaining, each with a position, stakes and influence | Whose bargaining produced this outcome, and who won? | Inter-ministerial contests over a scheme's design; "where you stand depends on where you sit" |
The line to use: Model I explains what the government says it decided; Models II and III explain what actually emerged. Deploying all three on an Indian example — the design of a centrally sponsored scheme, or a disaster response — converts a descriptive answer into an analytical one.
The Techniques
This is the half the syllabus names and most notes omit. Techniques divide into those that support an individual analyst and those that structure a group.
The Three Group Techniques Worth Knowing Precisely
- Brainstorming (Alex Osborn). Generate the maximum volume of ideas with judgement deliberately suspended; evaluation comes later. Its weakness is production blocking — only one person speaks at a time — and dominance by the confident.
- Nominal Group Technique (Delbecq and Van de Ven). Members first write ideas silently and independently, then present them round-robin without discussion, then discuss for clarification only, then rank privately. The silent phase is the point: it prevents the first speaker from anchoring the whole group.
- Delphi (developed at the RAND Corporation). Experts respond to structured questionnaires in successive rounds, never meeting, with a controlled summary of the previous round fed back before each new one. Anonymity removes status effects and dominance; iteration allows convergence without capitulation. It is the technique of choice for long-horizon forecasting where evidence is thin — technology assessment, epidemic projection, long-range demand estimation.
How Decisions Fail: Behavioural and Group Pathologies
- Groupthink (Irving Janis). Cohesive groups under pressure suppress dissent to preserve unanimity. Its symptoms are an illusion of invulnerability, collective rationalisation, belief in the group's inherent morality, stereotyping of opponents, direct pressure on doubters, self-censorship, an illusion of unanimity, and self-appointed mindguards who shield the group from disconfirming information. Janis's remedies are structural: assign a devil's advocate, invite outside experts, have the leader withhold their preference initially, and split into subgroups.
- Escalation of commitment. Continuing to invest in a failing course because of what has already been sunk. The Indian version is the project that must be completed because too much has been spent to abandon it.
- Cognitive biases (Kahneman and Tversky). Anchoring — the first number offered shapes all subsequent estimates, which is why last year's allocation determines this year's. Availability — vivid recent events dominate risk assessment, so the last disaster drives the next preparedness plan. Confirmation bias. Loss aversion, formalised in prospect theory. System 1 and System 2 — fast intuitive judgement against slow deliberate reasoning.
- Risk shift and group polarisation. Groups often reach positions more extreme than their members' initial private views.
- Abilene paradox. A group collectively decides on a course no individual member wants, because each mistakes the others' silence for agreement — a failure of communication rather than of preference.
- Wicked problems (Rittel and Webber, 1973). Some problems have no definitive formulation, no stopping rule and no true-or-false solution, only better-or-worse ones. Climate adaptation, urban air quality and manual scavenging are wicked in this precise sense, and applying structured optimisation to them is a category error.
Unanimity means every participant explicitly agrees; a single objection defeats the decision. Consensus means general agreement is recorded without a formal vote — dissent may exist but is not pressed, and the decision carries. The two are frequently conflated and are operationally very different: unanimity gives every member a veto; consensus gives the chair the power to read the sense of the meeting.
The Indian illustration is exact. The GST Council under Article 279A has a formal voting formula — a three-fourths weighted majority, with the Centre holding one-third and the states two-thirds of the vote. Yet it has operated almost entirely by consensus, with a formal vote taken only rarely. The design gives the Council a fallback; the practice avoids using it, because a vote would convert a cooperative forum into an adversarial one.
The evaluative point to close on. Consensus preserves relationships and protects the minority from visible defeat, but it can conceal coerced agreement and diffuse accountability — nobody is recorded as having decided. Unanimity guarantees genuine assent and invites paralysis. Follett's integration is the third possibility: not counting agreement but constructing a solution that meets the underlying interests of all sides.
Decision-Making and Paper 2: Indian Administration
| Concept | Paper 2 Chapter | Indian Instrument | Honest Critique |
|---|---|---|---|
| The choice phase institutionalised | Ch. 4 Union Government | The Cabinet Note, whose prescribed format requires alternatives considered and inter-ministerial comments to be set out; the Transaction of Business Rules; Cabinet Committees including CCS and CCEA; the Committee of Secretaries; Groups of Ministers | The Cabinet Note is a rational-model instrument on paper. In practice alternatives are frequently recorded after the preferred option has been settled — which is the 2023 narrowing point in its most consequential form. |
| Programmed decisions | Ch. 9 Financial Management; Ch. 8 Civil Services | General Financial Rules 2017 and the Delegation of Financial Powers Rules; the Central Secretariat Manual of Office Procedure; faceless assessment in income tax; rule-based DBT transfers; e-Office workflows | Programming raises consistency and removes the discretion that genuine hardship cases require. Faceless systems handle the standard case well and the exceptional case badly. |
| Non-programmed decisions | Ch. 7 District Administration; Ch. 14 Significant Issues | Disaster response under the DM Act 2005 and the incident response system; pandemic-era decisions on containment and relief; unprecedented industrial or ecological accidents | Simon's rule that higher levels face proportionately more non-programmed decisions implies analytical capacity should be concentrated there — yet India's senior levels remain burdened with routine file approval. |
| Appraisal techniques | Ch. 9 Financial Management | Cost-benefit appraisal in project clearance; the Public Investment Board and the Expenditure Finance Committee at prescribed thresholds; Detailed Project Reports; environmental and social impact appraisal under LARR 2013 | Appraisal is strong on monetised cost and weak on distributional consequence. Cost-effectiveness analysis is the appropriate tool where benefits resist pricing, and it is under-used. |
| Group technique in government | Ch. 4 Union Government; Ch. 5 Plans and Priorities | Expert committees and commissions; NITI Aayog consultations; Delphi-style expert elicitation in technology and epidemic forecasting; pre-legislative consultation | Committees are used to legitimise as often as to decide. The structured techniques that would actually improve group judgement — nominal group discipline, anonymity, formal devil's advocacy — are almost never used. |
| Consensus versus unanimity | Ch. 2 Constitutional Framework | The GST Council under Article 279A with its weighted formula but consensus practice; collective responsibility of the Council of Ministers under Article 75(3); the Inter-State Council | Consensus preserves the forum and conceals who conceded. Where voting weight is structurally unequal, consensus can be the polite name for the stronger party's position prevailing. |
| Risk aversion and the decision environment | Ch. 8 Civil Services; Ch. 14 Significant Issues | The 2018 amendment to the Prevention of Corruption Act and its Section 17A requirement of prior approval before investigating decisions taken in the discharge of official functions; T. S. R. Subramanian (2013) on written instructions | The most important Indian fact about decision-making is that not deciding is often the rational choice. Where the downside of a wrong decision is investigation and the upside of a right one is nothing in particular, delay is a rational response to the incentive structure. |
| Anchoring in budgeting | Ch. 9 Financial Management | Incremental budgeting where last year's allocation determines this year's; the persistence of line items long after their purpose has lapsed | Kahneman's anchoring and Lindblom's incrementalism describe the same phenomenon from psychology and politics respectively. Zero-based budgeting was the attempt to break the anchor, and it did not survive its own transaction cost. |
| Feedback into the next decision | Ch. 9 Financial Management; Ch. 10 Administrative Reforms | Outcome Budget and the Output-Outcome Monitoring Framework; the Development Monitoring and Evaluation Office; social audit; CAG performance audit; the PRAGATI review platform | India evaluates abundantly and closes the loop rarely. Evaluation findings seldom re-enter the intelligence phase of the next cycle, which breaks Simon's sequence at exactly the point where learning would occur. |
| Allison's Models II and III | Ch. 4 Union Government; Ch. 5 Plans and Priorities | Inter-ministerial contests over scheme design and allocation; standard operating procedures determining the shape of relief regardless of the disaster's character | Model I is how policy is announced; Models II and III are how it was produced. Writing all three on one Indian example is the fastest route to an analytical answer. |
| Behavioural design | Ch. 10 Administrative Reforms; Ch. 14 Significant Issues | The Economic Survey 2018-19 chapter on policy for homo sapiens rather than homo economicus; default-based enrolment in social security; framing in sanitation and nutrition communication | Nudge assumes the designer knows the citizen's true interest — a value premise presented as a factual one, and Waldo's objection in modern dress. |
Cross-Links Across Paper 1
| Topic | The Link To Make |
|---|---|
| Simon's decision-making theory (Ch. 2) | The theoretical foundation of this chapter. Bounded rationality, satisficing, the programmed–non-programmed taxonomy and decision premises all originate there. |
| Barnard (Ch. 2) | The art of executive decision lies substantially in not deciding — not prematurely, not what cannot be made effective, and not what others should decide. |
| Communication (Ch. 3) | Step 6 of the sequence. Information is the raw material of decision, and a decision not transmitted is not a decision. UPSC has linked the two directly. |
| Leadership (Ch. 3) | The Vroom–Yetton–Jago model prescribes how much participation a decision warrants, running from autocratic through consultative to group — the bridge between this topic and leadership theory. |
| Public Policy (Ch. 10) | Lindblom, Etzioni, Dror, the garbage can and Allison belong equally to policy formulation. Prepare once, deploy in both. |
| Techniques of Administrative Improvement (Ch. 11) | PERT, CPM, network analysis and MIS are named there and are decision techniques here. |
| Public Choice (Ch. 1) | Niskanen's budget-maximising bureau is a model of official decision-making driven by self-interest. |
| Accountability and Control (Ch. 5) | Audit and legislative scrutiny are feedback mechanisms; the fear they generate is itself an input into the next decision. |
UPSC Previous Year Questions — Verified
| Year | Question |
|---|---|
| 2023 | "The process of decisions … is largely a technique of narrowing choices." Explain. |
| 2008 | Examine the respective roles of facts and values in the decision-making process. Is it possible to make value-free decisions in the government system? How can government decisions be made more rational? |
| 1998 | "'Consensus' and 'Unanimity' are used as styles in decision-making." Comment. |
| 1990 | "A management which takes its environment as given … is pursuing a dangerous course." Comment on the environment of decision-making. |
| Also asked | "Information constitutes the life-blood of the functioning of an organisation." Explain the utility and significance of communication in decision-making. |
| Also asked | "While the entrepreneur makes strategic decisions, the manager makes tactical ones." Elucidate. |
| 2021 (Ch. 2 crossover) | "Two-dimensional taxonomy was used by Herbert Simon to describe the degree to which decisions are programmed or non-programmed." Explain. |
| Trend note | The stems are conceptual and quotation-led, not descriptive. None has asked "explain the steps in decision-making". Prepare to argue about the nature of deciding, not to recite a sequence. |
The stem warns that management which takes its environment as given pursues a dangerous course. Two readings, and a good answer gives both.
The systems reading. An organisation that treats the environment as fixed is behaving as a closed system — it will be outrun by changes it never registered. Katz and Kahn's point about coding applies: the system attends only to signals its formats capture, so environmental change outside those categories is invisible until it arrives as a crisis.
The strategic reading. The environment is not merely to be scanned but shaped. A regulator can alter the market it regulates; a state can change the incentives its citizens face. Treating conditions as given forfeits that capacity — and Barnard's third executive function, formulating purpose, presupposes an organisation that acts on its environment rather than only adapting to it.
The Indian illustration: a department that plans against last year's rainfall pattern, last year's migration flows or last year's technology is taking the environment as given. Chapter 3's answer is better intelligence; Chapter 4's answer is contingency and requisite variety.
Practice Questions For Test Series
- "Deciding is narrowing, and power lies with whoever controls the narrowing." Critically examine.
- "The syllabus says process and techniques." Assess the adequacy of quantitative techniques for public decisions under uncertainty.
- Apply Allison's three models to the design and implementation of a centrally sponsored scheme.
- "Consensus conceals who conceded; unanimity invites paralysis." Evaluate with reference to the GST Council.
- "In Indian administration, not deciding is frequently the rational choice." Examine the incentive structure that produces this.
- Compare Delphi, nominal group technique and brainstorming as remedies for the pathologies of group decision-making.
- "Anchoring and incrementalism describe the same phenomenon from psychology and politics." Discuss with reference to budgeting.
- "Wicked problems defeat the rational model by construction, not by difficulty." Comment.
Answer-Writing Toolkit
Two Introductions
- Foundational: "Simon's claim that decision-making is the heart of administration reoriented the discipline: administration is not primarily structure or rules but a continuous stream of choices, and everything else exists to shape them."
- Analytical: "A public decision differs from a private one in three respects — it is plural, since many contribute to what one person pronounces; it is exposed to pressures a firm never faces; and it cannot be indefinitely deferred, because a flood does not wait for better information."
Two Conclusions
- Balanced: "No single model captures public decision-making. The rational model supplies the benchmark, Simon the psychological realism, Lindblom the political realism, Etzioni the working synthesis, and Dror the warning that societies which must transform cannot rely on incremental adjustment. The techniques matter equally, and the examinable skill is matching technique to condition rather than reciting either list."
- Indian: "India's decision architecture is strong at intelligence and choice and thin at design and review. Alternatives are recorded rather than generated, evaluation rarely re-enters the next cycle, and an incentive structure in which a wrong decision invites investigation while a right one earns nothing makes delay rational. Reform must address the environment of decision, not merely its procedure."
Ten Power Keywords
- Intelligence, design, choice, review — Simon's phases.
- Programmed and non-programmed — with traditional and modern techniques.
- Satisficing and aspiration adaptation — the moving threshold.
- Successive limited comparisons — Lindblom's branch method.
- Partisan mutual adjustment — coordination without a coordinator.
- Mixed scanning — broad-angle camera plus zoom lens.
- Extra-rational elements — Dror's addition.
- Organised anarchy — the garbage can's precondition.
- Groupthink and mindguards — Janis.
- Anchoring and availability — Kahneman and Tversky.
Frequently Asked Questions
What are the phases of the decision-making process?
Simon identified intelligence, in which the environment is scanned and the problem identified; design, in which alternatives are invented and analysed; and choice, in which one course is selected. He later added review, in which past choices are assessed and their outcomes feed the next decision. In operational terms this expands into identifying the problem, gathering information, generating alternatives, evaluating them, selecting, communicating, implementing, and monitoring.
What is the difference between programmed and non-programmed decisions?
Programmed decisions are repetitive, routine and structured, with a definite procedure already established — sanctioning a pension, processing a routine claim. Non-programmed decisions are novel, unstructured and consequential, with no ready-made method — responding to an unprecedented disaster or designing a new policy. Higher organisational levels face proportionately more non-programmed decisions.
What are the main models of decision-making?
The rational-comprehensive model, which assumes complete information and optimisation; Simon's bounded rationality and satisficing; Lindblom's incrementalism or successive limited comparisons; Etzioni's mixed scanning, combining rational analysis for fundamental decisions with incremental analysis for detailed ones; Dror's normative-optimum model, adding extra-rational elements; and the garbage can model of Cohen, March and Olsen, in which problems, solutions and participants couple largely by chance in organised anarchies.
What is the Delphi technique?
Delphi is a structured method developed at the RAND Corporation for gathering expert judgement. Experts respond to questionnaires in successive rounds without ever meeting, and a controlled summary of the previous round is fed back before each new one. Anonymity removes status effects and dominance, while iteration allows opinion to converge without anyone having to concede publicly. It suits long-horizon forecasting where hard evidence is thin.
What is groupthink?
Groupthink is Irving Janis's term for the tendency of cohesive groups under pressure to suppress dissent in order to preserve unanimity. Its symptoms include an illusion of invulnerability, collective rationalisation, stereotyping of opponents, direct pressure on doubters, self-censorship, an illusion of unanimity, and self-appointed mindguards who shield the group from disconfirming information. Remedies include assigning a devil's advocate, inviting outside experts, and having the leader withhold their own preference initially.
What decision-making techniques are used in Indian administration?
Cost-benefit and cost-effectiveness appraisal in project clearance through the Public Investment Board and Expenditure Finance Committee; network analysis such as PERT and CPM in project scheduling; management information systems and dashboards; expert committees and consultations approximating group techniques; and increasingly data analytics and decision support systems. The Cabinet Note format, which requires alternatives and inter-ministerial comments to be recorded, is itself a structured decision technique.
Conclusion
Decision-making is where every other topic in this paper converges. Structure supplies the premises, communication carries the information, motivation determines the effort, leadership sets the participation, and accountability shapes the risk the decider is willing to bear. The syllabus is right to place it first in Administrative Behaviour.
Two conclusions are worth carrying. First, the 2023 insight: deciding is narrowing, which means the consequential act is often the filtering that happens before the file reaches the deciding authority — and therefore that power lies with whoever controls the shortlist. Second, the Indian diagnosis: the country's decision architecture generates plenty of intelligence and a functioning choice apparatus, but its design phase is thin, its feedback loop is broken, and its incentive structure makes delay rational. Techniques and models can improve a decision once it is being taken. Neither addresses an environment in which the safest decision is not to take one.
Key Takeaways
- Read the syllabus entry as two halves. "Process and techniques" — an answer that covers Simon and Lindblom but names no technique has answered half the question. Prepare CBA, decision trees, OR, game theory, payoff criteria and PERT alongside Delphi, nominal group technique and devil's advocacy.
- The 2023 stem is the one to master. Deciding is subtractive: the conceivable is cut to the legally permissible, then the financially feasible, then the administratively implementable, then the politically acceptable. Close with the critical turn — power lies with whoever controls the narrowing, which is Bachrach and Baratz's second face of power.
- Learn the models as a logical progression, each conceding what the one above assumed: rational-comprehensive → Simon removes full information → Lindblom removes the shared objective → garbage can removes even the problem-solution link. Dror is the most useful for Indian answers, because incrementalism suits societies that need not transform.
- Allison's three lenses are the fastest route to an analytical answer: Model I explains what government says it decided; Models II and III explain what actually emerged.
- Match technique to condition — certainty invites optimisation, risk invites expected value and decision trees, uncertainty invites Delphi, scenarios and judgement. Applying linear programming to a wicked problem is a category error, not a poor choice.
- Know the three group techniques precisely: brainstorming (volume, judgement suspended), nominal group (silent generation first, which stops the first speaker anchoring everyone), Delphi (anonymous expert rounds with controlled feedback, participants never meet).
- Groupthink needs its symptoms named — illusion of invulnerability, collective rationalisation, pressure on doubters, self-censorship, illusion of unanimity, and mindguards. Add escalation of commitment, anchoring, availability, and Rittel and Webber's wicked problems.
- For the 1998 stem: unanimity gives every member a veto; consensus gives the chair power to read the sense of the meeting. The GST Council has a weighted formula it almost never uses — designed for voting, operated by consensus.
- The sharpest Indian point: not deciding is frequently the rational choice. Where a wrong decision invites investigation and a right one earns nothing in particular, delay is a rational response to the incentive structure — which is why Section 17A of the PC Act belongs in this answer.
