Oligarchy Without Politburo: Measuring New Class Formation in Decentralized Autonomous Organizations A Djilas-Michels Framework and the Nomenklatura Index

Abstract

Decentralized autonomous organizations (DAOs) were supposed to prevent oligarchy. On the evidence, they haven't. Voting power is concentrated, delegation keeps producing superrepresentatives, and proposal turnout hovers below ten percent (Han et al. 2024; Fabrega et al. 2025). The empirical literature has measured this pattern many times over, but it has not had a theoretical frame equal to what it measures. This paper offers one. Drawing on Milovan Djilas's The New Class (1957) and Robert Michels's Political Parties (1911), I argue that DAO governance reproduces-not metaphorically, but structurally-the features of the communist nomenklatura: a self-perpetuating stratum that rules through control of the administrative apparatus rather than through ownership, that selects its members by cooptation rather than election, and that accrues privileges attached to position. I operationalize this through the Nomenklatura Index (ONI), a seven-dimension composite derived directly from Djilas's categories and piloted on three DAOs-Compound, MakerDAO/Sky, and Gitcoin-chosen to span the range of oligarchic formation I expected to see. The pilot does what it is supposed to do: ONI discriminates across the three cases in the predicted order and with meaningful spread. I close with design and regulatory implications. The governance-design ones (rotation mandates, supermajorities on self-dealing, deliberation minima) target the Djilasian mechanisms directly rather than working around them; the regulatory ones supply courts and agencies with a diagnostic vocabulary for the liability questions opened by CFTC v. Ooki DAO and unlikely to close anytime soon.