Why do states obey international rules when there is no world government to enforce them? Abbott, Keohane, Moravcsik, Slaughter, and Snidal (2000) tried to answer this with the concept of “legalization.” They argued that institutions can be compared by looking at three dimensions: obligation, precision, and delegation. In their view, a treaty with strong obligation, detailed rules, and a third-party enforcer is more “legalized” than one with vague promises and no monitoring. This framework gave international relations scholars a way to treat law as a variable, something that could be measured and tested.

But not everyone was convinced. Finnemore and Toope (2001) pushed back, saying that law cannot be reduced to these three dimensions. For them, law works through social meaning and practice, not just formal rules. Actors follow rules partly because they see them as legitimate or appropriate, not simply because obligation or delegation can be coded as “high” or “low.” Their critique was that legalization treats law as if it were mechanical — turn up obligation and compliance follows — when in fact law is closer to a shared language that shapes how states understand their own behavior.

This debate built on earlier regime theory. Stein (1982) showed that regimes help states solve problems of coordination and collaboration in an anarchic world, while Keohane and Martin (1995) argued that institutions provide information, focal points, and reassurance against cheating. Legalization was the next step: it added a more legal flavor to institutionalist theory, showing that some regimes were not just cooperative arrangements but also had the characteristics of law. Yet the very precision that made legalization appealing also risked flattening law into a checklist.

From today’s perspective, the question might be less whether legalization explains compliance and more whether the whole legalization debate reflects the discipline’s desire to measure what is easiest to code rather than what is most important. The danger is that by focusing on obligation and delegation, we miss the messy reality of how actors actually argue about law, reinterpret it, or strategically use its authority. Is legalization really a theory of law, or just a way for political scientists to make law fit their datasets?

Two extensions stand out. First, legalization can be understood not as a fixed property of institutions but as something actors use. States invoke “obligation” or “precision” when it strengthens their claims, while NGOs highlight the moral force of law even when enforcement is weak. Seen this way, legalization is a resource in political struggles, not just a variable. Second, the framework needs to move beyond states. Increasingly, private actors — from international standard-setting bodies to corporations — create rules that function like law. Do Abbott et al.’s dimensions capture these forms of authority, or do we need new ways of thinking about what counts as “legalized”?

The deeper question may be less about whether states comply with law, and more about why it matters to call something law in the first place. Legalization theory gave us useful tools, but the critiques remind us that the politics of defining law may be more revealing than the coding of its dimensions.