Countervailing Capital: Collective Voting Ownership and the Tail Index of Wealth

Proposals for collective ownership of corporate equity are usually assessed against the level of wealth inequality. This paper assesses one such proposal against the shape of the upper tail. For a top-wealth process with an additive income floor, multiplicative risk and an excess drift 𝜇 relative to aggregate growth, the stationary distribution is inverse-gamma with Pareto tail index 𝛼 = 1−2𝜇/𝜎2. Collective voting ownership reaches this index through one channel: it raises the marginal cost of extracting private benefits of control, and so lowers the control-rent component of 𝜇. Writing 𝜌 for that component, moving the index from its observed value 𝛼obs to a target 𝛼𝑡 requires 𝜌 > (𝜎2/2)(𝛼𝑡 − 𝛼obs) as a necessary condition, and a threshold larger by the factor 𝐻/(1 + 𝐻) as a sufficient one, where 𝐻 is the elasticity of extraction cost to outside voting power. The conditions are stated in estimable quantities and they fail for a recognisable class of economies. Calibrated with tail indices from Vermeulen (2018), top-wealth growth volatility from Gomez (2023) and control premia from Dyck and Zingales (2004) and Nenova (2003), the rent implied by observed premia exceeds the requirement for a finite-variance tail in eight of the nine countries where both quantities are estimated, and falls short in the United States by 0.63 percentage points per year. The instrument therefore has traction where control rents are large rather than where measured wealth concentration is highest. Two results limit the claim. Poisson turnover of controlling-owner status at a realistic rate raises the pooled tail index from 1.47 to 1.72 in simulation, so the aggregate tail responds to the controlling type’s rent by less than a fixed-weight mixture implies. And the governance block must be held by one fund per firm: spreading a given collective share across 𝑛 competing funds leaves none of them pivotal, so an architecture that partitions firms across funds is admissible where one that splits each firm is not.