According to the latest poll of polls by Real Clear Politics, Obama is ahead in Texas by roughly 2-3 percent and trails Clinton in Ohio by 7. If both states were to vote today, however, Obama would win Texas by a substantially larger margin and probably split Ohio with Clinton. Why? Two reasons: young voters and what I call "the cell phone effect."
Obama has outperformed the polling data in almost every contest on and since Super Tuesday. Heading into the Potomac Primaries, Obama registered 18 and 22 point leads, respectively, in the latest Virginia and Maryland poll. When the votes were counted, Obama had won Virginia by 29 and Maryland by 23. Obama defeated Clinton 64% to 35 in Virginia, overperforming the polls by 11 points, because he won the 18-29 vote 76% to 24. Obama defeated Clinton in Maryland 60% to 37, but only slightly overperformed the polls because he won young voters by a similar (but still lopsided) 64% to 33 margin. In other words, young voters don't seem as significant when its a multi-generational blowout - as it was in Maryland. However, young voters can and have to date turned otherwise close contests into big wins for Obama.
That's what happened in Wisconsin. Obama registered a 4 point lead in the polls. On election day, however, Obama thumped Clinton by 17. Obama won Wisconsin 58%-41, overperforming the polls by 13 points, in large part because he won the 18-29 vote 70%-26.
The inability of pollsters to measure Obama's momentum owes to consistent undersampling of young voters and oversampling of seniors, both of which occur because 14% of voters have cell phones but no landlines and roughly one-half of these voters are under the age of 30. Since young people are a core Obama constituency and seniors a core Clinton constituency, the cell phone effect translates into an unintentional but unmistakable pro-Clinton bias in the polls.
If the polling numbers hold, Obama will most likely deliver a knockout blow to Clinton in the Lone Star State.