Part 1: Summary Statistics
- President Obama job approval – last 30 days: 47-49% approve/disapprove (was 47-49% approve/disapprove)
- Generic Congressional Vote – last 30 days: 44-42% Republican (was 43-41% Republican)
- Congressional retirements (including defeats): 63 (52 in the House and 11 in the Senate)
- Congressional Primary defeats: 9 (8 in the House and 1 in the Senate)
Part 2: Projections (based on average of last 30 days of polling where states have held their primaries)
US Senate: 49 Democrats, 47 Republicans, 1 Independent, 3 Tossups (was 49-47-1 Democrat in last scorecard)
- Projected Republican gains in Nebraska and North Dakota
- Republican held seat in Indiana too close to call
- Democratic held seats in Montana and Virginia too close to call
- Republican held seat in Maine has an Independent in the race who is staying above 55% in polling
- (Note: 14 out of 33 Senate contests have not yet had their party primaries, so we are not in a position to call those races).
Governor: 30 Republicans, 19 Democrats, 1 Independent (was 30-19 Republican in last scorecard)
- Projected Republican gain in North Carolina
- (Note: 5 out of 11 Governor’s races do not have the party nominees selected yet, so we are not in a position to call those races).
Part 3: How we call a state – President, Governor, Senate
We believe that the best way to make our state by state calls in the Presidential race is to look at the actual percentage President Obama is receiving in the polls (and NOT the margin by which President Obama leads (or trails) Mitt Romney in any poll). There is a simple reason for this: when you’re talking about a controversial incumbent, those who are not explicitly supporting him in the polls will almost certainly end up voting for Romney on Election Day. Therefore, we have been compiling poll results by state and taking the average of that state’s poll results for the last 30 days. In the process of doing so, we have also considered Obama’s 2008 showing as a yardstick. Here are our criteria (which will be similar for statewide races for Governor or Senator):
(1) Safe Democratic/Obama or safe Republican/Romney (dark blue/red) – if no polls have been conducted within the last 30 days for that state, a 2008 Obama (or McCain) percentage of 60% or above gets this rating. If polls have been conducted, an Obama (or Romney) average of 50% or more gets this classification;
(2) Lean Democratic/Obama or lean Republican/Romney (light blue/red) – if no polls have been conducted within the last 30 days for that state, a 2008 Obama (or McCain) percentage between 53-59% gets this rating. If polls have been conducted, an Obama (or Romney) average percentage of 49% or less with a lead of 3 or more points will get this classification;
(3) Tossup (yellow) – if there was polling done, a candidate leads by less than 3 points or the 2008 election results had the winning candidate (Obama or McCain) receiving 52% or less;
(4) Senate/gubernatorial primary has not been held yet (or, no polls have been publicly released) (gray)
(5) No Senate/gubernatorial race in 2012 for this state (black)
Part 4: Presidential Scorecard (270 electoral votes required to win): Obama 271, Romney 206, Undecided 61 (prior scorecard: 271-191 Obama)
State by state polling has been relatively light this past week, although given that we are near the end of summer vacation and the Olympics are the main focus, this is not too surprising. The “macro” statistics of President Obama’s approval and the Generic Congressional vote have changed little, and we’re only making one change in our projections – North Carolina is moving from “tossup” to “leans Republican”. This is not a huge surprise, since (with the exception of 2008) North Carolina typically tilts Republican in Presidential contests. Furthermore, the Democratic enthusiasm/voter registration surge of 2008 that enabled Obama to carry the state are absent this year (our analysis of North Carolina voter registration was here )
We believe that voters will begin focusing on this race by the middle of the month, and THAT is when Romney has to make favorable impressions of voters with (1) his VP pick, and (2) his acceptance speech on August 30 (the Democrats hold their convention September 3-6). While it’s also important that Romney do well in the first debate, that event isn’t scheduled until October 3 – about the time early voting commences in several states.
Even though there has been heavy shelling from the Obama camp and/or surrogates (like Senate Majority Leader Harry Reid using hearsay evidence to accuse Romney of not paying his taxes ), the numbers haven’t really changed, except to prevent Romney from advancing much in the polls. Still, the Obama numbers aren’t that impressive at a statewide level if you consider the context of his 2008 showing.
In the “solid Obama” (dark blue) states that have conducted polling, Obama’s average share of the vote relative to 2008 has dropped from 60 to 52% – an 8% “plunge.” This dilution of support won’t affect Obama’s chances in these states (which collectively cast 195 electoral votes), but if this amount of “plunge” were to happen in more marginal states, the President would be in trouble. It’s also worth noting that in 8 of the 17 states in this category, Obama’s average poll number is between 49-51% – dangerously close to the “re-elect” line. Surprisingly, a recent poll in his state of Illinois shows him only with 51% of the vote (he carried the state with 62% four years ago).
Curiously, this “Obama plunge” is about the same in the “lean Obama” (light blue) states – in this category (worth 76 electoral votes), President Obama’s showing in the polls relative to 2008 has plunged from 55 to 48%. To put this in its proper context: Mitt Romney has taken a continued heavy pounding on the airwaves in these states, yet he is only 7 points behind an incumbent President. That is not an insurmountable poll deficit for Romney.
The situation is similarly dire for President Obama in the “tossup states”: his “plunge” in these states is from 53 to 47%, and Romney is averaging 44%. Obviously, some of these states (worth 61 electoral votes) will move into the “leans Republican” category at some point, as North Carolina just did.
Finally, in the states that are leaning or solidly Republican, the “Obama plunge” for states conducting polling is from 44 to 38%. Of course, these states are not likely to be contested anyway, and pollsters therefore are not likely to be doing much polling here.
Since we believe that most of those not explicitly for Obama will likely break towards Mitt Romney in the end, if we were to assume that the states where Obama is not polling at least 50% are Romney states, he (Romney) could receive up to 319 electoral votes (270 required to win). Does this contradict our scorecard showing Obama with more than the 270 electoral votes necessary to win? Actually, no. The scorecard looks at the lead Obama (a very known quantity) has over Mitt Romney (who is not a defined quantity in voters’ minds), and using that yardstick, Obama is (barely) across the finish line. But if you look at where Obama himself stands in the polls, his poll standing in 11 “lean Obama” or “solid Obama” states (totaling 122 electoral votes) is between 46-50%. Once voters start tuning in and/or pollsters change the way they survey voters, 46-50% is not a position where an incumbent President wants to be.
Part 5: Primary Results/Upcoming Events
Primary season has returned, as this past week, three contests (Texas runoff, Georgia primary, and Tennessee primary) were held, In general, incumbents fared well, although Republican primary voters again demonstrated that they are more interested in “read meat” candidates. In Tennessee freshman Chuck Fleischmann only received 39% of the primary vote (Tennessee does not have runoffs), while in Texas, Lt Governor David Dewhurst was upset 57-43% by Ted Cruz. This was another one of those races where the “young upstart” defeated an established, more moderate (in tone) candidate. Equally as impressive as the size of Cruz’s victory was the fact that 82% of those participating in the Texas runoff voted Republican. To put this figure in context, in 2010, Texas had a hotly contested governor’s race, and 69% of voters participated in the Republican contest. And in 2008, 68% of voters participated in the Democratic primary. This is further proof that Republican voters are more energized this year, and that energy will be worth an extra percentage point or two on Election Day.
Next week, we have primaries in Kansas, Michigan, Missouri, and Washington. There will be several hotly contested primaries, but what will be most interesting about next Tuesday’s races is that Washington (state) has nonpartisan primaries, and the results are typically a good predictor of which House seats will go Democratic or Republican in November. In fact, we had predicted from the August 2010 primary that an open seat was a GOP pickup in November, which in fact is what happened.
After those contests, 13 more states will have their primary contests before the middle of September, with Louisiana “bringing up the rear” with its primary on the same day in November as the Presidential election.