Oh hey there’s an election coming up

In contrast to 2008, I have not kept up with the daily changes and controversies of the presidential race. I have some thoughts about it but I prefer to take a more low key approach this time around.

I have enjoyed watching Real Clear Politics however and think they provide a nice service and up to date polling information. Unless something newsworthy comes up or I change my mind, I won’t be posting much until the election since  most other things will get lost until it is over.

 

  • Richard Willmer

    It’s been a fascinating few weeks, and it’s all so close.

    On the polling data: I’ve noticed what I regard as something of an anomoly: while the candidates are currently neck-and-neck in the national poll spread (with Obama having a slight ‘edge’ in the EC projections), the President’s Job Approval rating is currently showng a +2.6% spread, with an approval rating of about 50%. More curious still is that, when Gallup was showing Romeny at +5, it was, on the same day, showing Obama Job Approval at +7 (on 51%).

    Maybe I’m missing something here (I’m a simple soul at heart!), but this doesn’t seem to me to ‘match up’.

    It’s not so easy just now to tell who is the hare and who is the tortoise! I’ve bought my popcorn!

    I should also say at this point how saddened I am by the considerable loss of life, in both the Caribbean and the eastern United States, resulting from Super-storm Sandy. Over here, we were very struck by the fact that, at its zenith, it covered an area the size of western Europe.

  • Richard Willmer

    As of now, the President’s Job Approval ‘spread’ has inched up to +2.9% (at 50.0%). This is higher than that of President GW Bush eight years ago.

    But the Romney v. Obama popular support ‘spread’ remains unchanged at an effective ‘tie’.

  • F Young

    During this election cycle, Real Clear Politics has been consistently biased in favour of the Republican Party, as have been the Rasmussen and Gallup polls.

    For objective poll aggregation and commentary, check out Nate Silver’s fivethirtyeight.com or Sam Wang’s Princeton Election Consortium.

    http://fivethirtyeight.blogs.nytimes.com/

    http://election.princeton.edu/

    They show that the election has never been close since most of the bigger swing states have always supported Obama in polls. According to Silver, in order for Romney to win, there would need to be systematic bias in state polls (which he estimates at 16% probability).

    I am no expert, but what I have read suggests that there are now two main main ways that Romney could win: an unexpectedly higher turnout of Republicans or lower turnout of Democrats, and pro-Romney electoral fraud (rigged voting machines, voter challenges, vote rejection, inaccessible or closed polling stations). In most swing states, the election is being run by a Republican administration.

    It is unlikely that the result will be known on Nov 6. In a worst case scenario, it could take weeks.

  • Richard Willmer

    On the ‘raw percentages’ polling front, both the BBC and the HuffPost seem to show fairly similar situations.

    The analysis you show, F Young, looks very interesting at first glance. I’ll take a proper look later.

    As far as Rasmussen us concerned, it is perhaps the case that the methods (‘mechanically’-speaking) they use exclude disproportionately the poorest; it was Gallup that gave us that strange dissonance last week – Romney +5 (the ‘head-to-head’) and Obama +7 (job approval) at the same time.

  • Richard Willmer

    The Princeton analysis is very interesting. From what I can see, it suggests overall about a 2% ‘advantage’ to the President relative to what RCP is saying (which, nationally, is ‘Tie’ or thereabouts – as of today).

    My gut feeling is that this may well be about right, and for two reasons:-

    1. Rasmussen does much of its polling via the Internet, meaning that those without easy Internet access are under-represented. Hence, when Rasmussen says ‘Tie’ (which it has done for the last three days), ‘Obama +2′ might be nearer to the reality.

    2. The ‘Gallup dissonance’ between ‘head-to-head’ and ‘Obama job approval’ amounts to about ‘Obama +2′ (Romney +5 cf. Obama +7).

    @ F Young

    Does that sound pretty much ‘on the mark’ to you? (Obviously, I’m talking strictly numbers here; I have my own ‘preference’ when it comes to the winner, but feel that it would not be appropriate for me, as a Briton, albeit with American ancestry and connections, to go into that just now!) Looking at the numbers, that’s still IMHO ‘close’, but, given the state-by-state demographics, it looks like a second term for the President, ‘all other things being equal’.

  • F Young

    With respect to Rasmussen, a few issues have been identified by the poll analysts:

    1) they only poll landlines, while about one third of Americans only have cell phones, and they are disproportionately young,

    2) they only poll during a four hour window each day and make no attempt to reach people who do not answer the first call, which has the effect of excluding people who work at that time or are driving their kids around, who again tend to be younger.

    3) he adjust his results so a predetermined proportion identify themselves as R, D or I, which is a dubious practice that skews the results

    http://www.outsidethebeltway.com/rasmussens-2012-polling-has-had-a-republican-bias-all-year/

    By the way, analysts suspect that Rasmussen has started to artificially increase Obama’s popularity in the last couple of days to make Rasmussen’s polls match competing polls (which are much more professional), so that Rasmussen won’t look so bad after the election. If this is true, it would be a tacit admission that they don’t believe their own polls.

    The credibilty of all polls is reduced by the fact that fewer and fewer people now agree to give their opinion to pollsters, and they are unrepresentative. At best, only one tenth of people give their opinion; in many cases, it is more like 2 to 5%. This election may well decide whether electoral polling can even survive. If the polls are largely and unpredictably wrong (ie only a minuscule and unrepresentative sample of voters are willing to be polled; the inability to predict who will actually vote), the whole industry could disapear.

    http://www.electoral-vote.com/evp2012/Pres/Maps/Oct29.html#item-5

    Don’t be distracted by popular opinion. The media loves to hype how close national popular opinion is since it creates suspense and gives the illusion that all Americans can influence the election, which stimulates viewer and reader interest. In fact, the election will be decided only by swing voters who show up to vote in swing states, especially Ohio and Florida, and, I fear, by vote suppression and electoral fraud.

    http://www.electoral-vote.com/evp2012/Pres/Maps/Nov02.html#item-7

    http://www.electoral-vote.com/evp2012/Pres/Maps/Oct25.html#item-5

    http://www.electoral-vote.com/evp2012/Pres/Maps/Oct31.html#item-3

    There are reports that Obama has about 5000 lawyers lined up in Florida for election day, and 2500 in Ohio, to advocate on behalf of citizens who are obstructed at polling stations. Hundreds of lawyers will be flown in for recounts. Romney is apparently doing the same.

    http://www.electoral-vote.com/evp2012/Pres/Maps/Nov02.html#item-6

    http://www.electoral-vote.com/evp2012/Pres/Maps/Oct29.html#item-4

    The American electoral system is badly broken. It is outrageously penny wise and dollar foolish, and as corrupt as a third-world banana republic. If America invested in a uniform, professional system, it would be more credible, less corrupt and much cheaper to run when both internal and external costs are included.

  • Richard Willmer

    I can’t comment on the particular issues you raise, F Young, since my own knowledge of them is sparse.

    I do understand that the Electoral College system was perhaps designed for another era (when the Federal Government had fewer responsibilities than it does today in what is, amongst other things, a ‘nuclear age’ with high-speed systems of communication), and can see why many Americans might prefer a national, straightforward and clear ‘popular vote’ system. I also understand that, under the present system, ‘swing states’ get much more attention from the candidates than ‘safe’ or ‘likely’ ones. This seems rather unjust to residents of Texas, California, North Dakota, New York … : their issues are just as important as those of Floridians or North Carolinians (or even those living in Ohio – no disrespect, Warren!). On the other hand, a simple and national ‘popular vote’ system might compromise the influence of those living in small states, and could call possibly into question the genuinely federal nature of the Union.

    Maybe the solution would be to keep the Electoral College, but have ‘national standards’ on the conduct of general elections, plus – possibly – a small number of ‘federal electors’ that would cast their votes for the candidate with a majority / plurality of the national popular vote. But, in order to ‘honour’ the federal principal, the States would surely have to agree on how this might be done. Tricky?!

  • Richard Willmer

    Hope all goes well tomorrow, and the turnout is high! I’ll be thinking of you.

  • Richard Willmer

    It’s wasn’t so close after all!

    Now that all the votes have been counted, Obama comes out at just under +3.7% over Romney, with nearly 5 million more votes for the Democratic ticket. RCP said “Obama +0.7%”; how did the polls get so wrong? (RCP was 3% out in 2008 as well: they said Obama +5.5%; the results gave Obama +8.5%.)

    Obama won a whisker under 51% of the popular vote (Romney got 47.3%, others 1.7% between them). The win was not huge, but was decisive.


CLOSE | X

HIDE | X