An initial study investigating tolerance of group members who abuse a public good surprisingly showed that unselfish members (those who gave much toward the provision of the good but then used little of the good) were also targets for expulsion from the group. Two follow-up studies replicated this and ruled out explanations grounded in the target being seen as confused or unpredictable. A fourth study suggested that the target is seen by some as establishing an undesirable behavior standard and by others as a rule breaker. Individuals who formed either perception expressed a desire for the unselfish person to be removed from the group. …

The fact that generous people are unpopular is consistent with the well-documented aversion to exceptional individuals: dislike of those who seem extremely competent; displeasure with those who offer help; and, more recently, the rejection of those who adhere strongly to a moral position. …

Within a group task setting, social comparison tends to induce feelings of inter-personal competition. People feel driven to outdo the group member who is setting the standard. In a setting such as ours, the standard being set by the benevolent other is to give up a considerable amount of personal resources and receive only a small payoff in return. To compete with such a person means that one would need to give even more and take even less, not a very desirable prospect. Removal of this person would eliminate that competitive standard. Further, it is known that in social dilemma situations, people ignore the objective nature of their outcomes in favor of a subjective, relative evaluation of them. (more; HT Tyler)

Need any clearer evidence that status has a big relative (or positional) component?  Similar behavior happens in the real world.  Forager band hunters often exchange arrows, to hide who actually killed the animal everyone is eating; claiming credit for a kill is bad form and punished.  The book Managerial Dilemnas describes Hawthorne Works electrical factory workers in the 1930s:

The [official] incentive system was almost completely negated by social norms dictating roughly equal reported output.  The two pieces of equipment per day averaged out to 825 connections an hour by each of the wiremen.  Those who worked above this norm suffered some degree of social ostracism.  They were given nicknames such as “Speed King” and verbally chastised.  They were also likely to be the butt of a game known as “binging,” in which one person hit another as hard as possible on the upper arm, to which the other nominally had the right to respond by hitting back.  If a large number of people chose to play the binging game with the same norm violator, this amounted to a significant negative sanction. …

The four people who met or exceeded this norm were Krupa, Capek, Mueller, and Taylor. … Krupa, Capek, and Mueller were also the three most unpopular men in the room.  Taylor was highly popular, but managed his productivity without violating the norms, .. [as he] consistently reported making fewer connections than he actually did!  Because he underreported his outpout, his productivity could not be regarded as a greedy attenpt to get a larger share of the fixed revenue generated by the two pices of equipment produced per day; instead, his efforts could only be interpreted as a pure contribution to the public good. …

There were other people in the room who produced less than the norm.  They were tolerated with greater equanimity because their nonproductivity was not at such a level that it threatened the prescribed two unit per day goal, … [or as] “greediness.” Instead, they … had slightly less status, but were otherwise members in good standing. …

Members of the group were not allowed to “put on airs” by appearing to be better educated or better dressed than other members.  Such a distinction might presume to indicate that the peson was sufficiently exalted as to justify an unequal contribution to the cooperative work effort.  Indeed, any attempts by group members to seek such a socially distinct status were regarded by other members as almost as dangerous as free riding itself – and were santioned in the same spirit of tit-for-tat. (pp. 189-190)

Many suggest that even without a global agreement, the US should unilaterally tax carbon, as a contribution to the global public good, in order to shame other nations into following our example.  But might not this instead make such nations hate and retaliate against the US all the more?

Added Aug3: More from non-Western cultures:

In about half of these samples [we see] a phenomenon that is not observed beyond a trivial degree among typical undergraduate subjects: Many subjects engaged in anti-social punishment; that is, they paid to reduce the earnings of “overly” cooperative individuals (those who contributed more than the punisher did). The effect of this behavior on levels of cooperation was dramatic, completely compensating for the cooperation-inducing effects of punishment.