Sen. Barack Obama, D-Illinois, has proposed raising a number of taxes. Most notably perhaps he has discussed allowing the Bush income tax cuts (that Sen. John McCain, R-Ariz., voted against but now supports) to expire. He says that Americans earning less than $250,000 per household will see their taxes cut, not raised.
Mathias points out that 95.1% of the American people are in households that earn less than $200,000 -- so overwhelmingly most Americans will not see their income taxes increased, if Obama's math is correct.
(That's not to say that raising taxes on the top 4.9% -- which is, after all, disproportionately the group who make the jobs -- makes economic sense, necessarily. We're just looking at numerically how risky this position is in terms of how many Americans it will affect. And of course the wealthiest also happen to often be the most powerful as well, so this can be a group it's risky to irritate. Plus, Obama has proposed raising other taxes, including capital gains and divided taxes. As well as Social Security taxes -- though, again, only for those earning more than $250,000 a year.)
There's another component to this -- check out this map, by Mathias, which is interesting:
The designated areas are the top 50 major city areas by income, transplanted on an electoral map of the U.S. from 2004 featuring red Bush states and blue Kerry states. (The information about the major city areas came from the Tax Foundation.)
As you can see, most of these top 50 major city areas, income-wise are either in blue blue blue states (California, Massachusetts, New Jersey, or New York) or red red red states (Alaska, Texas), with fewer in swing states (with some notable exceptions such as Philadelphia, Penn., or West Palm Beach-Boca Raton, Florida.)
Mathias' conclusion: Obama's proposed income tax increases are not very risky at all, numerically speaking, in terms of number of voters or states he will impact. Overwhelmingly most Americans will not be touched, and the ones who will are largely in electorally decided states already.