We call it 'trick or treat' but we all know the chances are much higher of getting treats on Halloween night. Similarly, it seems that scientists have a higher probability of publishing research about reward rather than research about punishment.
I queried 'reward' and 'punishment' in the following databases of academic literature: Google Scholar, Scirus, Web of Science, SpringerLink, Ingenta, the journal Science, the journal Nature, and JSTOR. In the results below, you'll find 50% to 990% more publications with the keyword 'reward' than 'punishment', with the exception of the JSTOR database, which lists more articles with 'punishment' -- perhaps because it incorporates much more from the humanities and contains more historical research.
Why do scientists seem to be more focused on reward?
*Update November 12, 2011: Some scientists took beef (fairly) with this search, particularly because reward is used across disciplines in different fashions (think computer science) and because reward is both a verb and noun (while punishment is just a noun). To add a little more robustness, here is a similar search for 'altruism' and 'defection' that yields even more pronounced results (and more variable; check out Ingenta) in favor of academics favoring research related to reward more than punishment.