Tyler Cowen links to a paper that seeks to answer this question. What does it find?

In basketball, timeouts are believed to reverse the momentum of a game. However, here we show timeouts have no significant effect on the final outcomes of games. Moreover, we find that the timeout factor only appears to reinforce the game of dominant teams, meaning that only the most successful teams can find any positive benefit. We find no association with team payrolls, suggesting that richer teams are not particularly better at capitalizing on timeouts. Our findings support that strategic breaks have little impact on workplace performance and productivity.

This echoes the Freakonomics-cited research that disputes whether “momentum” or a “hot hand” actually exists in basketball.