Score contribution per author:
α: calibrated so average coauthorship-adjusted count equals average raw count
Neoclassical valuation methods often measure the contribution that non-market goods make to utility as income compensations. This circumvents Arrow's impossibility (AI) –a theoretical proof establishing the impossibility of social preferences – but those methods cannot be used in all settings. We build on Arrow's original proof, showing that with two additional axioms that allow for social learning, a second round of preference elicitation with a social announcement after the first, generates logically consistent social preferences. In short: deliberation leads to convergence. A ‘web-game’ aligning with this is trialed to select real world projects, in a deliberative way, with the board of an Australian Aboriginal Corporation. Analysis of the data collected in the trial validates our theory; our test for convergence is statistically significant at the 1% level. Our results also suggest complex social goods are relatively undervalued without deliberation. Most non-market valuation methods could be easily adapted to facilitate social learning.