When I was looking through Survey USA's latest 50-state Bush approval poll
, Hawaii again caught my eye as giving Bush an abnormally high job approval, considering its status as a "safe" blue state. To be sure, Hawaii has actually never given Bush a positive job approval
during Survey USA's nine-month run. However, according to Survey USA, it has pretty consistently given Bush his highest job approval among all Kerry states, even though there were quite a few stats that Kerry won by quite a lot less. In fact, there have even been some Bush states that have given Bush a lower net job approval than Hawaii these past several months, according to Survey USA.
Curious, I went through the last eight Survey USA 50-state polls, and compared Bush's national net job approval with his net job approval in Hawaii, adjusted for the partisan index. Over the past eight months, according to Survey USA, Bush's average net job approval nationwide has been -15.4. Since Hawaii has a pro-Democratic partisan index of 11.2, one would expect that Hawaii would deliver average net job approval ratings for Bush of around -26 or -27. However, Bush's average net approval in Hawaii has actually been only -11.1, a full 15.5 points off the expected approval numbers. If there were a pro-Republican shift of 15.5 points in Hawaii in the next electoral cycle, then Hawaii would move from a "safe Dem" partisan index of DNC +11.2, to a "lean Republican" partisan index of RNC +4.3. Good lord.
Growing even more curious, I went though the same process with 21 other states, all of which had a partisan index of less than 11.2 in favor of one party or the other. The vast majority came within a few points of their "expected" net job approval of Bush, but there were a few that stuck out:
State Average Approval Expected Approval Swing
AR -16.0 -8.1 DNC +7.9
OH -23.1 -15.7 DNC +7.5
MO -15.6 -10.6 DNC +5.0
NM -12.0 -17.1 RNC +5.1
MN -15.5 -21.3 RNC +5.8
HI -11.1 -26.6 RNC +15.5
Which would translate into:
State Current Partisan Index Swing New Partisan Index
AR RNC +7.3 DNC +7.9 DNC +0.6
OH DNC +0.3 DNC +7.5 DNC +7.8
MO RNC +4.7 DNC +5.0 DNC +0.3
NM DNC +1.7 RNC +5.1 RNC +3.4
MN DNC +5.9 RNC +5.8 DNC +0.1
HI DNC +11.2 RNC +15.5 RNC +4.3
In all of these states, if these swings were real, the differential would be enough to change the current status of the state on the electoral map. With a swing of 7.9 against Bush, Arkansas would no longer be "safe Republican," and instead become a "toss-up" state. Missouri would no longer be "lean Republican," and instead become a "toss-up". Minnesota would move from "lean Dem" to "toss-up" and New Mexico would move from "toss-up" to "lean Republican." Perhaps most interestingly, Ohio would no longer be a "toss-up" and would actually become safe
Dem. That would be pretty cool.
I have no idea if these long-term shifts that Survey USA is discovering will hold up in future months, or even if they are simply the result of having bad voter models. However, having consistent polling in all fifty states will potentially offer us early insight into electoral swings in 2006 and 2008. I, for one, would be perfectly happy if these swings were indeed real. The day Ohio becomes "safe Dem" is the day when Democrats are once again the natural ruling party in this country.