A Glance Forward to November’s Presidential Election

by Jackie_South on April 16, 2012

Now that the Republican nomination has been sewn up, if not confirmed, we are turning our attention to how the November presidential election itself will turn out.  As a reminder, the map below shows the 2008 result, with the numbers indicating the electoral college votes garnered by Barack Obama and John McCain.

Just to recap on the Republican race: with Mitt Romney’s victory in Wisconsin, over-turning a lead that Rick Santorum held there a month earlier, he has captured 659 delegates of the 1,144 he needs.  Santorum is in second place, on 275 delegates, and seeing that he couldn’t catch up, he has ‘suspended’ his campaign – code for throwing in the towel.  Newt Gingrich and Ron Paul both remain in the race, on 140 and 71 delegates respectively, but can only really act as an irritant and are unlikely to win any of the remaining 19 contests.

It looks a very tight race in November: today Gallup announced the start of their daily tracking poll lining up Romney as the candidate-presumptive against Obama.  The first poll puts Romney ahead on 47%, Obama two points behind on 45%.  This coincides with a daily approval poll that shows Obama having lost 2% in his approval rate to 45%.

This is part of the daily churn, and essentially really means that Obama and Romney are tied at the moment.  On the weekly aggregated approval rates, Obama has stayed in a range between 45% and 48% since the New Year, and currently has a weekly approval rate of 47% and a net approval over the last week of +1%.  Not over 50%, where he really needs to be to secure re-election, but better than the low 40’s rut he was stuck in through the latter half of 2011.

For a historical context, let’s compare that with all the post -war predecessors at the same point before the pending election:

  • George W Bush (Apr 2004): 52% (fell to 46% in May)
  • Bill Clinton (Apr 1996): 56%
  • George Bush Snr (Apr 1992): 42%
  • Ronald Reagan (Apr 1984): 52%
  • Jimmy Carter (Apr 1980): 39%
  • Gerald Ford (Apr 1976): 48%
  • Richard Nixon (Apr 1972): 54%
  • Lyndon Johnson (Apr 1968): 50% (fell to 46% in May)
  • Dwight Eisenhower (Apr 1956): 69%
  • Harry Truman (Apr 1948): 36%

So, Obama is in a healthier position than two presidents that went on to lose (Carter and Bush Snr) and one that went on to win (Truman) albeit in a very different age.  He is also improving slowly when George W Bush was declining in his ratings and is in a similar position to both him and Johnson in May of election year (most commentators believe that LBJ would have won in 1968 if he had stood).

Less good news is the comparison to Ford, who was at a similar level of approval in 1976 and went on to narrowly lose that November (Ford’s approval rate fell to 45% by June).  So, the over all message seems to be that Obama’s approval cannot afford to dip and that economic improvement needs to continue and earn him some credit.  That said, he does not need to improve that much and is heading in the right direction.

More optimistic is the polling at a state level.  I’ve aggregated the four most recent polls (if there are that many) since the start of the year to derive the map below.  This is a little statistically questionable, as I’ve mixed polls based on registered voters and those based on likely to vote voters, but this far out I think it is a viable approach.  The only weighting applied to the polls is sample size, so effectively a state with four polls is showing a single poll for the total of the four sample sizes.

The other caveat is that all these polls are against a generic Republican, as opposed to Romney by name, although of course the more recent polls would have respondees knowing that Romney was likely to be the candidate.

States in grey have yet to be polled this year. (source: Atlas of US Presidential Elections)

There are a few eye-brow raisers here.  First, Iowa shows a narrow Republican lead.  This seems instinctively wrong given Romney’s relative unpopularity in the Mid-West, the fact that Obama won the state by 10% in 2008 and that the state has voted Democrat in five of the last six elections (2004 was the odd year out).  This was a single poll in mid-February, giving a 2% lead with a 4% margin for error.

On the other hand, the polls feel a little generous to Obama in some other states: less than 5% behind in Arizona and only 6% behind in Tennessee both feel optimistic.  So do his leads in some states: 13% in Colorado particularly, but also 8% in New Hampshire, a state that looks eminently winnable by Romney.  6% and 7% leads in Nevada and Florida respectively also seem generous, and in the latter case there has been no polling since the Trayvon Martin issue blew up.

Georgia, however, looks genuinely interesting this time around.  Indiana, an Obama win in 2008, currently looks very likely to return to the Republican fold in 2012.

To win, a candidate needs to secure 270 of the 538 electoral college votes (EVs).  These polls give Obama 216 EVs in states where he leads by more than 10%.  To this you can add the 20 EVs from four states that have not been polled: Delaware, Hawaii, Maryland and Washington DC.  That means that to win, Obama needs to secure at least 34 EVs from elsewhere, and he leads by between 5-10% in six states with 90 EVs between them.

I’ve had a go at factoring those un-polled states in the final map.  As outlined above, four states are entered in as safe Obama states.  Of the rest, I have classified all as Republican wins of over 10%, except Montana which Obama only lost in 2008 by 2% (it won’t be as close this time, but I think will have a Republican lead of less than 10%).

All of this looks far better for Obama: like Bush in 2000, Obama may be able to lose the popular vote but still win the presidency.  The only thing that is certain is that this is an election with plenty more twists and turns before voters hit the ballot boxes.

Leave a Comment

{ 2 trackbacks }

Previous post:

Next post: