Is there any change in the contours of partisan support that has remained pretty much the same since 1996? It's a question I've been asking since the 2004 cycle began. (Here's an earlier look: http://www.usnews.com/usnews/opinion/baroneweb/mb_040511.htm.) It is unusual for the contours of partisan support to remain the same this long. Usually one party starts doing better in one region and worse in another. It happened between 1988 and 1996 when Democrats increased their standing in large (over 2 million) metro areas, especially in the seven very largest metro areas, while Republicans increased their strength, by a lesser amount, in the areas outside large metro areas. These countervailing trends left the electorate evenly divided, with 1 percent or less difference in the 2000 presidential elections and in the popular vote for the House in 1996, 1998, and 2000.
Browse through an archive of columns by Michael Barone.
To see whether there have been any changes, I looked at the 155 state polls taken in September and compiled in realclearpolitics.com on September 24. I calculated the average percentage margin in each state for George W. Bush or John Kerry and compared these percentage margins with those in the 2000 general election between Bush and Al Gore. All the numbers reported here have been rounded off; it makes no sense to dwell on tenths of a percent.
Three interesting patterns turned up.
First, Bush margins were up in 10 of realclearpolitics.com's 16 battleground states. Significant trends toward Democrats showed up only in Colorado (where most public polls in 2002 were Democratic, too) and West Virginia. Significant trends toward Bush appeared in several key states. Bush's margin was up 6 percent in Wisconsin; 5 percent in Pennsylvania; 4 percent in Florida, Iowa, and Maine; and 3 percent in Ohio and Missouri. The battleground seems to have shifted. Realclearpolitics.com has taken Arizona off its list of battleground states, and the Kerry campaign is off the air there and in Missouri as well. Among the 16 battleground states, Kerry leads by more than 2 percent in only two, Michigan and Washington.
Second, Bush's margins have held up well in most non-battleground Bush 2000 states. His margin is down 6 percent in North Carolina, 7 percent in South Dakota, 8 percent in Mississippi, and 11 percent in Idaho, but only North Carolina could possibly be in play, and Kerry-Edwards '04 is off the air there.
Third, and potentially the most important, the Democratic margin is down by more than 10 percent in three Gore states: 15 percent in New York, 14 percent in New Jersey, 12 percent in Maryland. These are all Metroliner states and states directly affected by the September 11 attacks. When George W. Bush arrived in New York for the Republican National Convention, he was greeted by a crowd of firefighters in Queens, and Bush strategists have been hoping for a surge of support from blue-collar voters in metro New York. These poll numbers suggest it may be happening.
Another factor that could be producing a weakening of Democrats in Metroliner states is Kerry's proposal to raise taxes on high earners. New Jersey is the No. 2 state in median household income; New York and Maryland have higher-than-average numbers of high-income voters. In 2000, when Bush's promise to cut taxes was met with skepticism, these people voted on cultural issues and mostly voted for Gore. It's possible that in 2004, when Kerry's promise to limit tax increases to those earning over $200,000 is being met with skepticism, some of them may choose to vote on economics and vote for Bush.