Here's the Stats Geek with a nice column on the Pirates' poor record this year (6-14) in games decided by one run.
The Stats Geek seems to place part of the blame on Jose Mesa. I'm no fan of Mesa, and he certainly deserves some part of the blame, but it wouldn't be fair to emphasize him too much. Mesa has pitched in nine games that ended up being decided by one run. In two of those, he gave up at least one run but the Pirates won by one run anyway. In three of them, he entered with a one-run lead and protected it. In the other four, he gave up runs and the Pirates lost.
If we subtract the seven games in which Mesa either protected a one-run lead or took the loss in a one-run game, the Pirates' record in one-run games is 3-10, which is even worse, percentage-wise, than the Bucs' 6-14 record in one-run games overall. Mesa isn't a good pitcher; it's bad if your team's closer isn't a good pitcher, and it's even worse when your team is overpaying that pitcher, but Mesa isn't the entire problem.
It probably isn't the Pirates' hitting, either. The Bucs have a .780 team OPS after the sixth inning, compared to .730 for their opponents; in "close and late" situations, they're at .779, compared to .651 for their opponents. So the problem doesn't seem to be 'clutch hitting.'
Actually, the problem may not be anything at all. As Studes points out in discussion of the Stats Geek article at Baseball Primer, the Pirates have a 9-5 record in two-run games, making them a more reasonable 15-19 in close games overall. A team's record in games decided by two runs is, for obvious reasons, less dependent upon luck than its record in games decided by one run, but focusing merely on one-run games leads one to believe the Pirates are a much better team than their record suggests, which isn't quite the case. Overall, the Pirates have scored 283 runs and allowed 299, giving them an expected won-loss record of 32-36, which is just one game better than their actual record. The Pirates' record in one-run games seems likely to even out (for reasons that are explained well by the Geek), but don't expect a big spike in their overall record when it does.