Domar & Musgrave (1944): Proportional Income Taxation and Risk-Taking
The 1944 QJE classic showing that a proportional tax leaves risk-taking undamaged — and can even encourage it — only with full loss offset: the state must share losses as fully as gains. Since no real tax system offers full refundability, this is the analytic hinge of the Schumpeterian objection to
Summary
Domar and Musgrave, "Proportional Income Taxation and Risk-Taking" (Quarterly Journal of Economics 58(3), 1944, 388–422), analyze how income taxation affects the willingness to hold risky rather than safe assets. The result that carries their names: with full loss offset — the government refunding tax value on losses as symmetrically as it collects on gains — a proportional tax makes the state a silent partner that shrinks both the upside and the downside of a gamble, leaving the risk-return trade-off intact; investors can restore their preferred risk exposure by scaling up, and taxation can even increase private risk-taking. Without loss offset, the asymmetry taxes wins at full rate while sharing losses only partially — discouraging risk.[1]
Why This Is the Hinge of the Rent-Tax Debate
The theoretical neutrality of rent-only corporate taxes — the allowance for corporate equity and cash-flow taxes — extends to risky investment only under the Domar–Musgrave condition (Bond & Devereux 1995 formalize this for business tax design).[2] Real systems fail it: loss carryforwards expire, earn no interest, and die with failed firms. So a real-world "rent" tax reaches the winners' realized quasi-rents at full rate while sharing the losers' losses incompletely — asymmetrically burdening exactly the high-variance ventures innovation consists of. This is the analytic core of the wiki's Schumpeterian objection: the objection holds against rent taxation as practiced even where theory answers it.
The condition is demanding but not utopian: Norway's post-2022 petroleum cash-flow tax pays cash refunds for the tax value of losses at a 78% combined rate — the existence proof that full refundability is administrable, at least in a sector where the rent is least contested (details and [VERIFY] flags on the cash-flow tax page).
See Also
- Objection: Taxing quasi-rents kills innovation — where this result does its work
- Allowance for Corporate Equity · Cash-Flow Tax
- Quasi-Rent — what asymmetric loss treatment ends up taxing
Sources
- Evsey Domar & Richard Musgrave (1944), "Proportional Income Taxation and Risk-Taking," Quarterly Journal of Economics 58(3), 388–422 — used for the loss-offset result (citation corroborated across QJE/RePEc/Semantic Scholar listings; the content characterization is the literature's standard reading — direct read proxy-blocked this session). OUP
- Stephen Bond & Michael Devereux (1995), "On the Design of a Neutral Business Tax under Uncertainty," Journal of Public Economics 58(1), 57–71 — used for the extension of rent-tax neutrality to uncertainty conditional on loss symmetry. RePEc