The contrast with Imperial China is the central comparison. China, unified under a single imperial administration, could suppress innovations deemed threatening to social order without facing competitive pressure from neighboring states of comparable power. When the Ming dynasty decided in the early fifteenth century to dismantle Zheng He's fleet and prohibit oceanic exploration, no competing Chinese state could exploit the withdrawal. The decision — not irrational within the logic of a unitary empire concerned with internal stability — was catastrophic for the continuation of Chinese commercial and technological leadership. It closed an entire domain of innovation without any external pressure to reverse the decision. European states, facing each other as competitors, could not make comparable withdrawals without handing advantage to rivals.
The mechanism operates in subtler ways as well. Religious dissidents persecuted in one European state could flee to another that welcomed them. This mobility constrained persecution and distributed the benefits of skilled migration. The Sephardic Jews expelled from Iberia in 1492 found refuge in Amsterdam, where their commercial networks contributed to the Dutch Golden Age. The Huguenots expelled from France in 1685 brought their textile expertise to the Netherlands, England, and Prussia, enriching their host societies at France's expense. These flows were not aberrations. They were the normal functioning of a competitive state system in which persecuting minorities meant losing them to competitors.
The same dynamic operated in intellectual and scientific domains. Galileo's conflict with the Catholic Church is remembered precisely because it was exceptional. The broader pattern was that heterodox thinkers could find institutional homes across a fragmented Europe. Descartes could move from France to the Netherlands. Spinoza could work in Amsterdam. The Royal Society of London, founded in 1660, created institutional protection for empirical inquiry that was copied by academies across Europe. No single political or religious authority could suppress scientific inquiry across the continent, which meant inquiry could accumulate across centuries despite episodes of local suppression.
Applied to the AI moment, competitive pluralism is partially present and trending toward concentration. The global technology market is competitive at the state level — the United States, China, and the European Union represent meaningfully different regulatory and commercial environments, and capital and talent flow among them. But within each of these jurisdictions, AI development is concentrating in a small number of firms with the capital, data, and computational resources required to train frontier models. The concentration of AI means is structurally concerning: if the competitive dynamics that once operated among many commercial actors collapse into competition among a handful of frontier labs, the incentive for institutional openness that competitive pluralism provides weakens. The broligarchs critique gains analytical weight.
The question for sustained AI bloom is whether the institutional openness that competitive pluralism produces can be preserved through antitrust enforcement, open-source AI development, and international regulatory arbitrage. If the AI industry consolidates into oligopoly, the structural incentive for institutional openness weakens. Firms with monopoly positions do not need to tolerate heterodox ideas or attract skilled migrants. They can extract rents from their position without facing competitive pressure. The loss of competitive pluralism at the industry level is one of the most concerning structural trends of the current moment, and it occurs silently — not through dramatic political events but through the quiet consolidation of capability in the firms that can afford frontier model development.
The concept builds on a long tradition in comparative political economy, including Eric Jones's The European Miracle (1981) and the work of Charles Tilly on state formation. Goldstone's contribution is to integrate competitive pluralism with the other institutional features of the Why Europe? framework, demonstrating that no single feature was sufficient and that the combination was historically rare.
Fragmentation as advantage. Apparent inefficiency of multiple competing jurisdictions produces structural incentives for openness that unitary authorities cannot replicate.
Exit as constraint. Skilled migrants can flee persecution; capital can flow to friendlier jurisdictions; innovations can be adopted by rival states. These exit options constrain persecutors.
Contrast with imperial unity. Song China's unitary administration could suppress disruptive innovation; no comparable suppression was possible in fragmented Europe.
AI-era application. The mechanism still operates at the state level among the US, China, and EU; it is weakening at the industry level as AI development concentrates.