Complex vs. Complicated

Apr 1

Written By Emika Otsuka

Idea level

Concept

Definition

Complex vs. Complicated is an ontological distinction that Yuen Yuen Ang draws between systems made up of interconnected elements that adapt to one another (complex—like trees), and mechanical objects composed of separate parts whose operations are linear and controllable (complicated—like toasters).

This distinction forms the starting point of Ang’s Adaptive Political Economy by correcting a fundamental classification error: treating complex social systems as complicated objects.

Once this distinction is made, it has downstream consequences for reorienting our understanding of core social dimensions, including causality, indeterminacy, human agency, and institutional design—implications that are typically overlooked in existing definitions of complexity.

Sources

First articulation:

Ang, Y.Y. (2016). How China Escaped the Poverty Trap. Introduction; Chapter 2: Directed Improvisation.

Theoretical synthesis:

Ang, Y.Y. (2024). “Adaptive Political Economy: Toward a New Paradigm.” World Politics. https://doi.org/10.1353/wp.2025.a954433. | [Open]

Public communication:

Ang, Y.Y. Complex doesn’t mean complicated—and what it Means for Political Economy.” Interview by Emily Babson. World Politics, 24 Feb 2026.

Genealogy

[Paradigm] Industrial–Colonial Paradigm
→ [Pillar] Mechanical thinking (treating societies as machines) → [Concept] Complicated (conflated with complex)
→ [Concept] Causality: linear / Human agency: control / Indeterminacy: risk / Designs: institutions as fixed rules
→ [Theory] Linear theories stuck in Chicken-and-Egg Fallacy of Development

Contrast with

[Paradigm] AIM (Adaptive, Inclusive, Moral) Political Economy
→ [Pillar] Adaptive Political Economy (APE)
→ [Concept] Complex (≠ complicated)
→ [Concept] Causality: non-linear / Human agency: influence / Indeterminacy: influence / Designs: meta-institutions as structures that enable adaptation and learning
→ [Model] Directed Improvisation: one example of meta-institution → [Theory] Coevolutionary Development → [Method] Mapping Coevolution

Quotes

[Ontological distinction] The terms ‘complicated’ and ‘complex’ are often conflated in daily language, but in fact they describe two completely different worlds… To study complicated worlds, we can parse out the different parts into separate categories of cause and effect... Much of our analyses have proceeded as if social worlds are complicated… Complex systems comprise many moving parts that interact with one another and change together, triggering outcomes that cannot be precisely controlled or predicted in advance… Political economies are complex.

— Ang, How China Escaped the Poverty Trap (2016), pp. 10.

[Starting point of paradigm construction] I adopt a paradigm that is different from the one that we currently embrace. Our conventional paradigm (the way we view the world) assumes a complicated—rather than complex—reality… Yet we all know that social worlds are not complicated; they are almost always complex.

Fortunately, just as we don’t always have to kill insects in order to study natural habitats, we don’t have to reduce complexity in order to make sense of complex worlds.

— Ang, How China Escaped the Poverty Trap (2016), pp. 10-11.

[Correcting classification error] The industrial paradigm is flawed because it contradicts social reality. Modern social science was born in an age of mass industrialization. That context has a profound, invisible influence on our ontology. We lack basic concepts to recognize systems that are not like toasters. They are perceived as nuisances, complications that stand in the way of advancing “parsimonious” social science theories.  

Thus, in APE, I begin by underscoring a classification error. It is an error to treat societies as complicated, because they are inherently complex, with mutually adaptive elements producing outcomes that may be influenced, but not controlled.

Making this ontological distinction is the first step toward a paradigm shift, because if the subject (like development) is misclassified, more errors will follow. 

— Ang, “Complex Doesn’t Mean Complicated” (2026).

[Downstream implications for social analyses] “The differences between complicated machines and complex systems are not semantic but have profound implications for the way social scientists understand causality, indeterminacy, human agency, and institutional design.”

— Ang, “Adaptive Political Economy” (2024).

Concept Constellation

Across Ang’s work, Complex vs. Complicated consistently co-appears with the following

concepts and analytic themes.