http://www.dailykos.com/story/2008/10/12/1601/5665/599/627909The Washington Post maintains a webpage titled "Political Landscape 2008." This webpage has been updated recently enough to reflect Obama's 13.8% polling lead in Pennsylvania, a trend that uses polls released as recently as Wednesday. On this webpage, the Washington Post collects recent polling data, and then declares a state to either be a "battleground," or leaning toward one party or the other. Here are some of their polling averages and diagnoses:
Obama +13.8%: Battleground state (PA)
Obama +10.4%: Battleground state (NH)
Obama +10.0%: Battleground state (NJ)
Obama +9.5%: Battleground state (IA)
Obama +9.0%: Battleground state (OR)
Obama +8.2%: Battleground state (MN)
Obama +8.2%: Battleground state (MI)
Obama +8.8%: Battleground state (WI)
Obama +7.3%: Battleground state (NM)
McCain +6.8%: Leaning Republican (GA)
Obama +5.1%: Battleground state (VA)
Obama +4.0%: Battleground state (CO)
McCain +3.8%: Leaning Republican (IN)
Obama +3.5%: Battleground state (OH)
Obama +3.1%: Battleground state (FL)
Obama +3.0%: Battleground state (NV)
McCain +2.2%: Leaning Republican (WV)
Notice anything wrong with this list? Could it perhaps be that any state where McCain leads, no matter his margin, is defined as "Leaning Republican?" Could it be that states where Obama leads by 7.3%-13.8% are defined as "battleground states," while states where McCain leads by 2.2%-6.8% are defined as "leaning Republican." Does the uneven math in this strike anyone as problematic?
The Washington Post claims that a 2.2% lead for McCain is larger than a 13.8% lead for Obama. That is objectively wrong and quantifiably unfair. This is as blatantly imbalanced as election reporting can possibly get. For more on this, see my related post on the persistent inability for big media to admit that Obama is over 270 electoral votes, no matter what polling indicates.
http://www.openleft.com/showDiary.do?diaryId=9018