Kerry On Top
Don’t Believe the Tripe
By Paul Rosenberg, Senior Editor

     For months now the conventional wisdom has been that the presidential race is essentially tied. But as usual, the conventional wisdom has been wrong, according to analysts who look at three major indicators of where the race is heading. In reality, Kerry has been ahead since April, widening his lead considerably before highly publicized lies about his service in Vietnam recently narrowed the race. Without understanding the lead Kerry had built, it is difficult to appreciate the Bush campaign’s need for such scurrilous attacks.
     The first expert indicator is approval of the incumbent, measured in three different ways. The second is state-by-state analysis of the swing states where the election is being fought. The third is cumulative analysis of nationwide polls, and where they are trending. All three indicators decisively showed Kerry winning through mid-August—a position he may well recapture once the GOP convention bounce wears off.
     As long ago as May 9, pollster John Zobgy wrote a piece, “The Election Is Kerry’s To Lose.” Zogby cited evidence of the first and third types. Citing his own mid-April poll, he noted Kerry was leading 47 to 44 percent in a two-way race, while “only 44 percent feel that the country is headed in the right direction and only 43 percent believe that President Bush deserves to be re-elected.” The “right direction” and “re-elect” numbers are two of the three ways of measuring approval of the incumbent—the two most stable and most reliable. The third—straight-forward “presidential approval”—jumps around a lot more, and is usually less indicative of solid support.
     But throughout the summer, the strongest evidence of Kerry’s lead came from examining the swing states—a task made easier by a handful of websites focused on tracking swing states or electoral votes (EVs) in general. Even after the Swift Boat attacks, on the weekend of August 28-29, two electoral vote sites—Federal Review and Election Projection—had Kerry up by around 90 EVs, 316-222 and 311-227 respectively, while a third, Electoral Vote Predictor, had Kerry 270—259. All are updated when new polls come out, and have maps distinguishing the strength of statewide leads. Because they use different methodologies, and differ in sensitivity to the latest polls, it’s best to consider all three. None show Bush ahead. Swing State Project updates EV totals less often, and has Kerry ahead 281-257, but looks much more closely at swing state races.
     Kerry built substantial momentum in expected swing states like Pennsylvania, Oregon and Washington, where Bush’s chances have grown increasingly remote, while forcing Bush to fight hard for states he might have assumed were secure. Kerry has led in Florida more often than not, put Virginia, North Carolina and Arkansas in play, and pulled ahead in Tennessee. That’s five of 11 Southern states where Bush faces a real fight. The Mountain West is weak as well, with Kerry holding onto New Mexico, and threatening in Colorado, Arizona and Nevada. With weakness in both his base regions, Bush has been much more on the defensive than the media has let on. A 30-point lead in Idaho or a 40-point lead in Utah carries very little weight just across the state line.
     The Zogby battleground poll released after mid-August showed Kerry ahead in all but two states, Ohio and West Virginia. He led in eight Gore states and six Bush states from 2000. Zogby also announced he would be expanding his battleground polls to include four formerly “safe” Bush states— Arizona, Colorado, North Carolina and Virginia. Virginia last went Democratic for LBJ in 1964.
     Because of the Electoral College, state-by-state analysis is the gold standard for understanding the election. Still, with substantial national media coverage, and over 7 million voters overseas, it would be foolish to ignore other approaches.
     As already noted, Zobgy said Bush was in trouble back in early May. But Zogby was not the first to see that Bush was in serious trouble. The first clear indication that Bush was in trouble goes back to the period of his seemingly greatest triumph—the “successful” invasion of Iraq. At the time, analyst Ruy Teixeira pointed out that, despite a 20-point jump in Bush’s approval rating, his re-elect number went up just 4 points from before the war—and, more importantly, remained below 50 percent. With a re-elect number that low, Teixeira felt confident that the more volatile approval numbers would come down—and they have.
     Bush’s approval ratings have spiked three times—after 9/11 they went around 90 percent, rising from the low 50s. After the invasion of Iraq they went up to around 75 percent, rising from the mid 50s. After the capture of Saddam Hussein they rose again, from around 50 percent to the high 50s. Each peak has been lower than the one before, and is followed by the same inexorable decline. While “right direction” and “re-elect” numbers have long been signaling trouble, Bush’s presidential approval numbers have been mostly between 40 and 50 percent since the beginning of May. When combined, these three indicators of incumbent performance clearly show Bush in deep trouble. He cannot be re-elected because people approve of him. He can only be re-elected if he makes Kerry even less appealing. So far, he has spent over $100 million trying to do just that—and he’s failed. Even the Swift Boat lies have fallen short. Kerry’s disapproval numbers remain well below Bush’s.
     Our final approach is cumulative analysis of polls. All polls have margins of error (MOEs) that make them somewhat unreliable. If one poll has Kerry leading 48-46 with an MOE of 3, the poll is statistically a tie. However, if 50 polls added together show Kerry leading 48-46, that difference is significant. Pollster Stan Greenberg of Democracy Corps showed Kerry leading Bush by 0.2 percent in April, 2.8 percent in May, 1.5 percent in June, 3.1 percent in July and 5 percent through the first half of August—all in two-candidate races. Including Nader reduces Kerry’s margin, but Nader’s percentage is about half what it was in May.
     There is one final twist: state level cumulative poll analysis. In mid-August, analyst Chris Bowers, of Swing State Project and MyDD.com, pointed out that Bush was below 50 percent average in twelve states he carried in 2000:
     “In each of these twelve states, he has been below fifty in a significant majority of all post-Super Tuesday polls. Further, in each of these twelve states, his raw score in the last five trial heats has averaged below fifty. These twelve states are worth a total of 136 electoral votes, leaving Bush 142 seemingly safe electoral votes.”
     While the Republican Convention should give Bush a brief, modest bounce, the months-long trends for Kerry are clear signs of solid strength that should re-emerge by late September. It is Kerry’s strength in these fundamentals that guarantees a nasty campaign coming from the Republicans. There is no other way that they can win.