Altruism does exist, and not just because people want the "good feeling" of having done "good deeds."
I've seen Levitt of the University of Chicago speak several times about an altruism experiment that economists conducted showing the lack of altruism in people. In response, one of the audience members once sent him $20 in an envelope and said, "Explain this."
While that's a facetious example, there's plenty of examples of more self-less altruism towards strangers. I think plenty of research shows the altruistic motive towards families, which is partly motivated by evolution.
But altruism towards strangers?
I think an interesting experiment to conduct would be a double blind experiment regarding altruism. Put someone in a situation where they can help, hurt, or ignore someone in need, and there is no cost to the subject of picking any of the three options. But they get no reward for doing the good deed, and no one is watching.
(Of course this experiment is sort of impossible by design, cuz the experimenter has to be watching)
How many people choose to do the good deed without social affirmation of their good deed?
Maybe I'll reconsider pursuing that PhD in Economics.