On August 15, 2022, FIFA suspended the All India Football Federation within hours of a Supreme Court order appointing an administrative committee to reform the federation's governance. The same FIFA spent seven years cooperating with Swiss prosecutors investigating its own leadership without directing any sanction at the Swiss Football Association. This paper argues that the asymmetry reflects a four-century conceptual mechanism: the non-justiciability of political decisions. FIFA extracted this mechanism from constitutional law, privatized it, and redeployed it as a selective enforcement tool against states with high football identity and low institutional power relative to the organization. Drawing on Ernst Fraenkel's analysis of the dual state (El Estado dual, Trotta, 2022), the Baker v. Carr genealogy of the political question doctrine, and Extended Phenotype Theory (EPT), I construct a three-layer model of what I term the FIFA Immunity Memeplex. I apply a two-variable analytical model to eight documented cases across three continents, demonstrating that the non-interference principle is activated with a selectivity that FIFA's own official justifications cannot explain. The paper proposes a three-zone taxonomy of 'non-justiciable football matters' that distinguishes a narrow legitimate core from a broad zone that constitutional theory and EPT both identify as properly subject to state oversight. Argentine constitutional jurisprudence provides an additional analytical lever: the Supreme Court of Justice of the Nation progressively dismantled its own non-justiciability enclave from Cullen v. Llerena (1893) through Fayt, Bussi, and Brussa, consistently reducing the immune zone when fundamental rights or institutional power were at stake. A private Swiss organization's statutory claim to equivalent immunity is structurally weaker, not stronger, than the constitutional doctrine the Argentine court chose to constrain. Argentina's ongoing criminal prosecution of federation officials (2026) provides a real-time test of the model's predictive power.
Ignacio Adrian LERER (Tue,) studied this question.