The Soft State
Myrdal introduced the concept of the "soft state" in Asian Drama to describe institutional formations characterized by weak enforcement capacity, pervasive corruption, the capture of public policy by elite interests, and systematic gaps between formal rules and practiced reality. Soft states possess formal institutions — constitutions, regulatory frameworks, legal codes — but lack the enforcement infrastructure that would give those institutions binding force. The formal architecture exists; the operational architecture does not. The result is a governance environment in which rules apply to the weak and dissolve for the powerful, producing cumulative patterns of inequality that the formal institutions are ostensibly designed to prevent.
In the AI Story
The concept was developed to explain why post-war development aid to South and Southeast Asia had failed to produce the economic transformation its designers had promised. Capital, technology, and policy advice had been transferred in substantial quantities. The institutional environments receiving these transfers had absorbed them into existing patterns of elite capture, rendering the transfers mechanisms of advantage consolidation rather than broad-based development. Myrdal's conclusion was stark: technology transfer without institutional strengthening produces dependency, not development.
The transferability of the concept to AI governance is exact. Contemporary AI regulation — including the most comprehensive frameworks such as the EU AI Act — operates in institutional environments that combine formal regulatory architecture with weak enforcement capacity. The disclosure requirements exist. The risk classifications exist. The safety standards exist. Whether they bind the conduct of the companies they regulate depends on enforcement — which depends on staffing, technical capacity, political will, and resistance to capture by the industries being regulated. Across every jurisdiction, enforcement capacity lags formal authority by orders of magnitude.
The soft state concept also illuminates the international dimension of AI governance. Nations with weak regulatory capacity become sites where AI systems are deployed under conditions the developers would not accept in their home jurisdictions — a pattern familiar from environmental, labor, and pharmaceutical regulation. The asymmetry is not accidental; it is the exploitation of institutional softness as a competitive advantage in deployment strategy.
For Myrdal, the soft state was not a cultural characteristic or a pathology of particular regions but a structural formation produced by specific political-economic conditions. It could be overcome through deliberate institutional strengthening — public administration reform, enforcement capacity building, anti-corruption infrastructure, democratic deepening — but the strengthening required sustained political commitment and was actively resisted by the elite interests benefiting from institutional softness.
Origin
Myrdal developed the concept most fully in Asian Drama (1968) and elaborated it in The Challenge of World Poverty (1970). The analysis drew on his extensive fieldwork across South and Southeast Asia, where he observed the pattern repeatedly across contexts that differed in culture, history, and formal political arrangement but shared the underlying structural dynamics of enforcement weakness and elite capture.
Key Ideas
Formal architecture without operational capacity. Institutions exist on paper; they do not exist in enforcement.
Selective binding. Rules apply to the weak and dissolve for the powerful, producing systematic inequality in regulatory incidence.
Technology transfer absorption. Transfers into soft states are absorbed into existing patterns of elite advantage rather than disrupting them.
International regulatory arbitrage. Soft institutional environments become deployment sites for practices developers would not accept at home.
Structural, not cultural. Soft states are products of specific political-economic conditions, not intrinsic features of particular societies.
Appears in the Orange Pill Cycle
Further reading
- Gunnar Myrdal, Asian Drama, vol. II, ch. 18 (Pantheon, 1968)
- Gunnar Myrdal, The Challenge of World Poverty (Pantheon, 1970)
- Francis Fukuyama, State-Building (Cornell, 2004)