Score contribution per author:
α: calibrated so average coauthorship-adjusted count equals average raw count
We provide a new proof for the optimality of deductible insurance that does not depend on the expected-utility hypothesis. Our model uses only first- and second-degree stochastic dominance arguments.