I realize this is just what one model says, this week may represent Clinton's high water mark at the peak of the convention bounce, and blah blah blah. But currently Nate Silver projects that Hillary Clinton has a better chance of winning Texas (24.6%) than Donald Trump has at winning Pennsylvania (19.3%)
(Note the above numbers come if you have the models set to the "polls only" forecast. It also works on the "now-cast" setting, although the percentages are different. But if you set it to "polls-plus" PA is better for Trump, with 24.8% chance and TX is a lot worse for Clinton at 10.1%. So really 2 out of Silver's 3 models suggest that Clinton has a better chance winning Texas than Trump does winning Pennsylvania)
(Note the above numbers come if you have the models set to the "polls only" forecast. It also works on the "now-cast" setting, although the percentages are different. But if you set it to "polls-plus" PA is better for Trump, with 24.8% chance and TX is a lot worse for Clinton at 10.1%. So really 2 out of Silver's 3 models suggest that Clinton has a better chance winning Texas than Trump does winning Pennsylvania)