One of the key political terms in which variable definitions cause a lot of confusion is “battleground states.” In the context of a presidential general election, the term ultimately means states that the two major party candidates target with time and money, particularly as election day nears. In earlier stages of the presidential cycle, the term is more speculative, as reflected in the inveterate promises of this or that candidate to “expand the battleground” by placing more states into play. Early general election polling–particularly now that state-by-state polls have become more common–is often the source of such speculation; that’s why you hear talk from the Obama campaign that the Democrat will at least throw a scare into McCain in such previously invulnerable GOP bastions as Georgia or North Carolina, and why McCainiacs like to tout outlier polls showing him well ahead in PA or competitive in CA.
Over at OpenLeft, Chris Bowers has made an effort to identify the likely 2008 battleground states by a more scientific method: using the close 2004 race as the baseline, and adjusting the numbers to reflect recent demographic trends, with the final ranking also reflecting polling evidence and common sense. Chris also takes the simple but often forgotten step of stipulating a very close general election, not just because he thinks that’s likely, but because the very term “battleground” becomes misleading in an electoral blowout. That’s why he assigns such states as Florida and Pennsylvania–almost invariably described as battleground states in most lists–to McCain and Obama, respectively, on the theory that if McCain’s leading in PA or Obama’s leading in FL down the homestretch, the candidate is probably going to be well ahead nationally. (He also views IA and MN as Obama states in a close race, and WV and MO as McCain states in a close race). This doesn’t mean the campaigns will ignore such states in the general election; it is, instead, a prediction of where the big effort will and won’t be made in the final push.
Chris comes up with a list of eight likely battleground states, with five (CO, NH, OH, NM and WI) leaning to Obama, and three (MI, NV and VA) leaning to McCain.
You can quibble with his list, which Chris acknowleges as very preliminary, but it is based on a sound methodology that generally avoids the pro-Obama temptations of overinterpreting positive 2006 election results or making outlandish assumptions about turnout, and the pro-McCain temptation of giving the Arizonan a thumb on the scales in every state where Obama did poorly in this year’s primaries.
One important point that Chris doesn’t explicitly address, but that we might as well get used to, is the recurring possibility of an election in which the popular vote winner loses the electoral vote, and thus the election. It obviously happened (with an assist from the Supreme Court) to the benefit of Republicans in 2000, and could well have happened to the benefit of Democrats in 2004 (in the latter case because Kerry’s percentage margin of defeat in states like OH, IA and NM was a lot smaller than his overall popular vote deficit). Given the likelihood that Obama will run better in solid red states than is usual for a Democrat, you’d have to guess–and it is just a guess–that he’s the more likely victim of this sort of miscarriage of political justice. But you wage election campaigns with the system you have, and those who fear an electoral vote/popular vote split better get behind the state-based National Popular Vote initiative (which would reward EVs to the national popular vote winner) pronto.