Via orgtheory, an interesting graph showing the evolution of the average number of references in the American Economic Review and the American Journal of Sociology since the 1950s:
If anyone has time to waste, it would be interesting to see the trend in political science journals; my guess is that it is somewhere between the two.
PoliSciZurich 
I’d be curious about citations per page – AJS is known for its willingness to publish very lengthy pieces – it’s one of the few journals that do – while AER has probably increased the number of very short papers in the “papers and proceedings” section.
True. And a random example suggests that AJS articles were indeed much shorter in the 1960s. Interesting.
A comment by an anonymous reviewer who recommended rejection of a paper: “This manuscript does not cite my work.”
If you submitted to AJS you have no excuse!
If!