The rich, as F Scott Fitzgerald said, are different from you and me.

Just how different is revealed in a paper by Florida State University's James Ang, Mississippi State University's Gregory Nagel and Jun Yang from Indiana University.

Basically, they find evidence that how much the CEO gets paid depends very much on his social circles. CEOs attend various social functions, they meet up at exclusive golf clubs and country clubs and that inevitably creates some social ranking.

Put simply, the higher up and better connected you are, the more you get.

In their paper, Is there a Social Circle Premium in CEO Compensation?, they write:

"Regardless of whether CEOs are actively engaged in social comparisons, they are aware of and are concerned about their rankings in social circles. As a result, the utility/happiness of a CEO depends not only on his own wealth, but also on his wealth relative to the (subjective) benchmark in his social circle. If the benchmark is expected to run up, it is reasonable for the CEO to demand more to stay equally happy. It is also rational for the board to satisfy the CEO's demand on pay raise if other boards are doing so."

Their research covered the period from 1993 to 2005. Executive pay data was drawn from the ExecuComp database. Information on social elites was drawn from the zip codes on the 2004 Social Register and Forbes provided details of their educational background. They also looked at memberships of golf clubs.

The researchers discovered a "social circle premium" to CEO pay. They found that the average compensation of CEOs in a social circle increases with the size of the social circle and that excess pay (actual pay in excess of the economically predicted pay) increases with pay ranks and the size of the social circle. Pay raises were higher for CEOs ranked higher in their social circles. The bigger the social circle, the more this was likely. In other words, CEO pay increases when the CEO moves up the social ladders. Indeed, the compensation of CEOs in the top deciles of their social circles was 18 times more sensitive to the number of peer CEOs than CEOs in the bottom 50th percentile of their social circles. Also golfing contacts with peer CEOs were strongly associated with higher CEO compensation ($132 thousand for each additional golfing contact).

So are CEOs highly paid because of their connections? Or does their exorbitant pay open doors and extend their social circle? As the writers suggest, it's a chicken and egg argument but in the end, it's pretty much the same result:

"The results support the view that CEOs are concerned about their ranking in the social circle. Whether actively engaged in social status comparison or not, they are aware of their social status. This leads to the impetus to demand more pay, knowing that other CEOs in the social circle will be doing the same."