Three Levels Of Analysis In International Relations

How the Three Levels Actually Work When You're Doing the Analysis

Most people learn the Three Levels Of Analysis In International Relations from a textbook and immediately assume they're mutually exclusive frameworks. They aren't. In practice, you bounce between them constantly, and the skill is knowing when one explanation is falling apart and you need to shift gears. Kenneth Waltz laid this out in Man, the State, and War, and J. David Singer later formalized it into the three-level model. But formalizing it doesn't mean it stays neat. Real cases are messy, and the levels bleed into each other in ways that will annoy you if you're not expecting it.

Individual Level: Leaders and Their Heads

This is the level most people default to because it's the easiest to narrative-drive. You look at a decision and ask: what does this specific person believe, what's their psychology, what experiences shaped them? Putin's actions in Ukraine get a lot of individual-level analysis. His KGB background, his sense of historical grievance about the Soviet collapse, his apparent willingness to gamble on NATO's cohesion fracturing. That matters. It explains choices that structural accounts struggle with on their own. The problem is you can never prove the counterfactual. Did Russia act this way because of Putin, or would it have acted similarly under a different leader given the same structural conditions? You'll read confident arguments on both sides, and neither is conclusively provable with available data.

State Level: Domestic Politics and Institutions

When individual-level explanations start feeling thin, you push down to the state level. Here you're looking at institutions, interest groups, public opinion, economic structures, bureaucratic politics. The question becomes: what domestic forces are shaping what this country can actually do? I spent a chunk of 2023 analyzing why certain European states were slower to supply certain weapons to Ukraine than others. On the individual level, you could point to leader personality. But the real driver was institutional: which countries had procurement systems with legitimate veto points, which had defense industries structured around export markets that created competing priorities, which had political coalitions that would fracture over military aid commitments. The individual leaders mattered less than I initially thought. The common pitfall here is confusing correlation with causation on the domestic side. You see a country with a strong military-industrial complex and assume that complex drove its foreign policy toward intervention. Sometimes it did. Sometimes the intervention happened for reasons completely unrelated, and the industrial complex just got dragged along afterward. Always check the timeline.

System Level: Anarchy and Power Distribution

This is the level Waltz himself argued deserved the most attention. You abstract away from leaders and domestic politics entirely and look at the international system: how many great powers exist, how is power distributed, what are the structural incentives created by anarchy. The system-level explanation for the Ukraine conflict is straightforward: Russia perceives a unipolar system that threatens its great power status, NATO expansion compresses its strategic depth, and in an anarchic system states prioritize survival and relative gains. This explains the broad shape of the conflict without saying anything about Putin's personality or Germany's domestic politics. System-level analysis has a real weakness that beginners often miss: it tends to predict outcomes that are too generic. Yes, balancing behavior is likely when a power perceives a threat. But which country balances, how, when, and against what exactly — the system level often can't specify. That's where you need to pull back up to the state or individual level for precision.

How to Actually Use This Without Getting Confused

Start with the system level if you're trying to understand broad patterns. It gives you the baseline expectations. Then push down to the state level to explain variation among states facing similar structural pressures. Then hit the individual level if you still have unexplained cases. The reverse approach works too, especially when you're analyzing a specific decision. Start with the individual actor, then ask what domestic constraints they're operating under, then ask whether those constraints are unique to this case or reflect deeper structural pressures. Here's something beginners consistently get wrong: treating the levels as competing explanations rather than complementary ones. They're not. The most useful analysis uses all three simultaneously, showing how they interact. A system-level pressure creates the conditions. State-level politics determines which tools are available. Individual-level decisions determine when and how those tools get used.

One edge case that trips people up: when a leader's personal beliefs align perfectly with what the structure would predict anyway. You see Russia acting aggressively and it looks like either Putin's personality or systemic pressure is driving it. In those moments the framework stops being helpful unless you introduce a timeline. Did the structural incentives exist before the leader took power? Did policy shift after leadership changed in ways that don't track with structural change? Those mismatches are where you find evidence of which level is actually doing the work. The Three Levels Of Analysis In International Relations isn't a method that produces a single correct answer. It's a set of lenses. Used well, it makes your analysis more honest about what it can and can't explain. Used poorly, it becomes an excuse for picking whichever level produces the conclusion you already wanted. I've read far too many papers where the author claims multi-level analysis but really just cherry-picks whichever level supports their argument and ignores the rest. It's also worth noting that this framework was built primarily for state-centric analysis of high politics. It struggles with non-state actors, transnational movements, economic interdependence that cuts across state boundaries, and issues like climate change where the causal chains run through every level simultaneously and can't be cleanly separated. When those cases come up, you either extend the framework uncomfortably or acknowledge that it doesn't have much to say.

The workaround I use is to treat the levels as a checklist rather than a hierarchy. When I'm evaluating any foreign policy decision, I explicitly note which level provides the strongest explanation and which provides the weakest, then flag where the gaps are. That forces me to admit uncertainty instead of masking it with whichever level tells the most satisfying story.