Foreign-policy analysis: levels of analysis & bureaucratic politics

From systems to decision-makers

Foreign-policy analysis (FPA) asks not what states want but why a particular state chose a particular act at a particular moment. It opens the black box that realist systemic theory deliberately keeps shut. Its founding device is the levels-of-analysis problem, framed by Kenneth Waltz in Man, the State, and War (1959) as three images of the causes of war, and sharpened by J. David Singer in his 1961 World Politics article 'The Level-of-Analysis Problem in International Relations'.

Waltz's three images

• First image (individual/human nature): war originates in the psychology, ambition or misperception of leaders. Robert Jervis's Perception and Misperception in International Politics (1976) is the canonical first-image study; cognitive biases, analogies (the 'Munich' analogy of 1938) and emotion drive choice.
• Second image (state/domestic): the internal constitution of the state—regime type, ideology, public opinion, interest groups—explains behaviour. Democratic peace theory (Doyle, 1983) and the 'second-image reversed' (Gourevitch, 1978) live here.
• Third image (international system): anarchy and the distribution of capabilities compel behaviour regardless of who governs. This is the structural-realist level Waltz himself privileged in Theory of International Politics (1979).

Waltz's claim is that the first two images locate the efficient causes while the third supplies the permissive cause: anarchy allows war, but does not by itself explain any single war. Singer's contribution was methodological—warning that mixing levels produces incoherent explanation, and that the analyst must choose a level deliberately and stick to it.

Why a fourth level matters

Later FPA added a decision-making level beneath the state: the bureaucracies, committees and standard operating procedures through which policy is actually produced. Richard Snyder, H.W. Bruck and Burton Sapin pioneered this in Decision-Making as an Approach to the Study of International Politics (1954), insisting that 'the state' is a fiction and that only identifiable officials decide. This decision-making turn is what distinguishes FPA from grand IR theory: it treats the policy output as the product of human beings operating inside organisations under stress, deadline and incomplete information.

The practical payoff is diagnostic. Confronted with any case—India's 1998 Pokhran-II tests, the US 2003 invasion of Iraq, China's 2013 launch of the Belt and Road Initiative—the analyst can ask: is this driven by a leader's worldview (first), by domestic coalitions and regime imperatives (second), by systemic pressure and the balance of power (third), or by organisational routine and bureaucratic bargaining (fourth)? A complete answer usually layers all four, but a disciplined answer names the dominant level and justifies the choice.

Allison's three models

The decision-making level acquired its definitive vocabulary from Graham Allison's Essence of Decision: Explaining the Cuban Missile Crisis (1971; revised with Philip Zelikow, 1999). Analysing the October 1962 crisis, Allison offered three competing 'conceptual lenses', each generating a different explanation of the same events.

• Model I — Rational Actor Model (RAM): the state is a unitary, value-maximising actor selecting the option with the best cost-benefit ratio. The blockade was the optimal middle course between inaction and air strike. This is the default assumption of realism and of most exam answers—and Allison's point is that it is one lens, not the truth.
• Model II — Organizational Process (Behaviour) Model: outputs are not choices but the product of pre-existing standard operating procedures (SOPs) of large organisations. Why were Soviet missiles in Cuba deployed in patterns identical to those in the USSR, betraying their presence? Because Soviet engineers followed home SOPs. Why could the US Navy not implement the blockade exactly as President Kennedy wished? Because the Navy executed its doctrine.
• Model III — Governmental (Bureaucratic) Politics Model: policy is the resultant of bargaining among players positioned across government, each pursuing the interest of their agency and their own stakes. Miles's Law—'Where you stand depends on where you sit'—captures it. The outcome is a political resultant, not anyone's preferred design.

Bureaucratic politics in practice

The bureaucratic-politics model explains pathologies that the rational-actor model cannot: turf wars, the dominance of 'options' framed to favour the framer, the persistence of policy after its rationale expires, and the gap between presidential intent and field implementation. Morton Halperin's Bureaucratic Politics and Foreign Policy (1974) extended the framework to peacetime weapons procurement and treaty negotiation.

Group-level pathologies complete the toolkit. Irving Janis's Victims of Groupthink (1972) explained the 1961 Bay of Pigs fiasco as the suppression of dissent inside a cohesive in-group; the same Kennedy circle, restructured to encourage dissent, performed better in 1962. Bounded rationality (Herbert Simon) and prospect theory (Kahneman and Tversky, 1979)—the finding that decision-makers are risk-acceptant over losses—now anchor first-image FPA.

High-yield critiques

Stephen Krasner's 1972 Foreign Policy article 'Are Bureaucracies Important? (Or Allison Wonderland)' is the standard counter: it argues the model lets leaders off the hook and overstates bureaucratic autonomy where presidents in fact set the agenda. For exams, the disciplined position is that Models II and III supplement rather than replace Model I—they explain the friction between intent and output.

Why this matters for the exam

FPA is the analytical hinge that lets you convert factual knowledge of any foreign-policy event into a structured answer—precisely what distinguishes a topper's script from a current-affairs summary.

UPSC GS-2. The IR syllabus demands analysis of 'India and its neighbourhood' and 'bilateral, regional and global groupings'. Examiners reward candidates who explain why a decision was taken, not merely what happened. Deploying the levels-of-analysis frame—was the 2019 abrogation of Article 370's external dimension driven by domestic politics (second image) or strategic signalling to Pakistan (third)?—demonstrates conceptual command. The rational-actor versus bureaucratic-politics distinction is directly usable on questions about why India's policy outputs sometimes lag stated intent.

FSOT. The Job Knowledge Test examines US government structure and the inter-agency process. Allison's Essence of Decision is effectively required reading; the Cuban Missile Crisis (October 1962), the National Security Council system, and the State–Defense–CIA bargaining dynamic are recurrent. The Structured Interview's group exercise literally simulates bureaucratic bargaining—knowing Miles's Law and groupthink is operationally useful, not merely academic.

CSS and BCS. The International Relations and Current Affairs papers set essay questions on decision-making and foreign-policy formulation. A candidate who can name Waltz's three images, Singer (1961), Allison's three models (1971), Janis on groupthink (1972) and Krasner's critique (1972) writes with an authority that scores in the top band.

The PYQ angle

Direct theory questions on FPA are rarer than on realism or liberalism, so the framework is most valuable as an answer architecture deployed on applied questions. When a paper asks 'Analyse the determinants of India's foreign policy' or 'Discuss the factors shaping US policy toward China', the levels-of-analysis structure—systemic, domestic, individual, bureaucratic—gives an instant, defensible skeleton.

High-yield retention list

• Waltz's three images: individual, state, system (1959).
• Singer, 'The Level-of-Analysis Problem' (1961): choose one level, do not mix.
• Snyder–Bruck–Sapin (1954): the decision-making turn.
• Allison's three models (1971/1999): Rational Actor, Organizational Process, Governmental Politics.
• Miles's Law: 'Where you stand depends on where you sit.'
• Janis, groupthink (1972): Bay of Pigs 1961 vs the restructured 1962 ExComm.
• Krasner (1972): the critique that bureaucracies are overstated.
• Jervis (1976) and prospect theory (1979): the cognitive first image.

Memorise the authors, dates and the one-line claim of each; deploy them as evidence, never as decoration.