As noted, Obama trounces Clinton among voters who care most about change, and she beats him as broadly among those more focused on experience. But another measure fleshes out this equation, and raises a potential vulnerability for Obama.
Two-thirds in both states say Clinton, if elected, would "do enough" to bring needed change to Washington; fewer, 56 percent in Ohio and 53 percent in Texas, say Obama has the kind of experience it takes to serve effectively as president.
Obama still has the majority's endorsement on experience — but at a considerably lower level than Clinton's acceptability on change. (In only a few groups does Obama fall short of a majority on experience, such as seniors and Hispanics in Texas.)
Additionally, while Obama leads as the most electable in November, more than six in 10 likely voters in both states say either candidate could beat John McCain, the front-running Republican.
In Texas, seniors and Hispanics say by 2-1 that only Clinton could beat McCain; African-Americans, by 3-1, say only Obama could do so.
In Ohio, college graduates, independents and blacks pick only Obama by especially wide margins.
Enthusiasm for the candidates — potentially a factor in turnout — is running about equal for Clinton and Obama, albeit a bit lower in Ohio than in Texas.
That's especially true for Clinton; in Texas 65 percent of her supporters describe themselves as "very enthusiastic" about supporting her; in Ohio, 53 percent. (Obama's numbers are 62 percent in Texas, 56 percent in Ohio.)
Turnout, naturally, is crucial. Clinton leads in Texas and Ohio alike among people who say they voted in the 2004 primary; new voters are better for Obama. Given their sharp differences, the relative mix of Hispanics and blacks voting in Texas is equally critical; Clinton's support may rely on a boost in turnout by Hispanics over 2004, when they accounted for 24 percent of voters.
On the other hand, Clinton arguably could prevail without a big Hispanic turnout, if instead seniors showed up in large numbers; they accounted for a sizable 26 percent of voters in the Texas Democratic primary in 2000, but then dropped to 19 percent in 2004.
Likely voters in this poll account for 24 percent of the adult population in Texas and 30 percent in Ohio. While actual turnout at those levels is unlikely, vote preference results are similar in likely voter models positing much lower turnout.