From weakest-link to best-shot: Correction

Here X is the socially available amount of the public good in a community of I individuals, while xi represents the contribution of individual i (i = 1, ..., I). In this Weakest-link situation, for two individuals A and B, the respective preference maps take on drastically different shapes northwest of the 450 line (where xA xB). In particular, A will never contemplate operating outside the northwest region, while B will always choose to be in the southeast region. I should have noticed that A's indifference curves must be strictly vertical in his relevant region, while B's must be strictly horizontal in his relevant region. (A degree of curvature does exist, in each case, but only in the nonrelevant region.) The diagram here, showing only the indifference curves in the relevant regions, is a corrected version of Figure 2 in the original paper. In the corrected diagram there is a best vertical indifference curve for individual A, UA, associated with the maximum xA that A would ever be willing to contribute regardless of how much greater B's production may be. Consequently, A's Reaction Curve RA runs from the origin to point F along the 450 line, and thenceforth along the vertical UA. For B similarly, his Reaction Curve RB runs from the origin to point G on the 450 line, and thenceforth along his best indifference curve the horizontal UB. Consistent with the argument of the original paper, the equilibrium will be at the upper limit of the range where the Reaction Curves overlie one another along the 450 line point F. The range between points F and G is like a 'contract curve' in that one party can gain only at the expense of