If you think about policy recommendations as attempts to move resources, there’s a straightforward reason why actors with sufficiently unaligned preferences might prefer to send “care more” messages over “care less” ones.
Let’s say you have coalition A, which contains potential “care more” target B, and does not contain potential “care less” target C. Then you have everything in neither C nor A, which we can call D. The resources freed up by a “care less” message about C are divided in some proportion between A and D, while a “care more”message about B benefits A exclusively. This dilution effect means that even if agents are constrained to sending true messages, if they can prioritize selfishly, they’re often going to favor “care more” over “care less” at similar levels of effectiveness.
If nearly all careabouting territory is defended, the dynamics are somewhat different, since “care more” mostly stops working. If a dominant coalition owns most of the resources, then the main source of variation is the occasional competing claim, and there’s a strong incentive for the large player to shut down smaller ones as they pop up. The very large coalition stands to reap the majority of the gains from “care less” attacks against outsiders. Cf. monotheism. On the other hand, small players have a comparatively strong shared incentive to attack the largest one, since even a fairly small shrinkage of its resource claims in percentage terms may free up what is to them quite a lot, for ensuing “care more” claims.
I found this slightly hard to parse, would be interested in someone writing this again… maybe just in slightly different words, maybe with real examples instead of A/B/C/D.
If you think about policy recommendations as attempts to move resources, there’s a straightforward reason why actors with sufficiently unaligned preferences might prefer to send “care more” messages over “care less” ones.
Let’s say you have coalition A, which contains potential “care more” target B, and does not contain potential “care less” target C. Then you have everything in neither C nor A, which we can call D. The resources freed up by a “care less” message about C are divided in some proportion between A and D, while a “care more”message about B benefits A exclusively. This dilution effect means that even if agents are constrained to sending true messages, if they can prioritize selfishly, they’re often going to favor “care more” over “care less” at similar levels of effectiveness.
If nearly all careabouting territory is defended, the dynamics are somewhat different, since “care more” mostly stops working. If a dominant coalition owns most of the resources, then the main source of variation is the occasional competing claim, and there’s a strong incentive for the large player to shut down smaller ones as they pop up. The very large coalition stands to reap the majority of the gains from “care less” attacks against outsiders. Cf. monotheism. On the other hand, small players have a comparatively strong shared incentive to attack the largest one, since even a fairly small shrinkage of its resource claims in percentage terms may free up what is to them quite a lot, for ensuing “care more” claims.
I found this slightly hard to parse, would be interested in someone writing this again… maybe just in slightly different words, maybe with real examples instead of A/B/C/D.
This is a rigorous version of my intuitive dissatisfaction with the OP.