And the winner is …

By Lorenzo

Political science models. President Obama’s re-election has already been thoroughly mythologised, particularly by those disappointed by said re-election. One can read worried, or admiring, examinations of the data-mining and thoroughly internetted campaign techniques used by the Obama campaign, complaints about media bias, dark musings about the implications for the future of the US and so forth.

Or one can read some of the political science commentary which will direct you to things such as this graph (via), which maps share of two major party vote against per capita (income) growth in the second and third quarters of the election year (i.e. the two quarters before the election) discounted by length of tenure of the incumbent Party.  As one can see, the remarkable thing about President Obama is how little either of his election wins is an “outlier” result.

As political scientist Larry Bartels observes in the post:

Perhaps the president deserves some sort of award for his unmatched fidelity to the “laws” (irony intended) of political science.

I was fairly confident that President Obama was going to be re-elected, because I read the Monkey Cage, a group political science blog in which the aforementioned post recently appeared. The blog had periodic posts on what the political science models were reporting on the US Presidential election prospects, and those models put the President’s re-election at a very high probability.

The US Presidential election is a gift for political modelling because it happens at the same time of the year every four years. It turns out, that one can get (by the standards of social science) remarkably accurate “fits” of the data with remarkably few variables. As political scientist John Sides observes in an earlier post on The Monkey Cage, this still leaves a role for political campaigns. But the next time you read any commentary on the election result — particularly commentary that implies President Obama’s re-election was a surprising result that needs special explanation — consider the above graph, and the similar results other political science models generated.

This is also the Saturday chit-chat post.


  1. kvd
    Posted January 12, 2013 at 9:56 am | Permalink

    There orta be a law against prepending a perfectly serviceable word such as ‘science’ with anything as naff as claimed in the post.

    And if a Masters in Public Administration, and a Masters in Communication leads a person seeking election to a position amounting to that of Shire Clerk to speak like this, then we need to urgently reassess the worth of some sectors of higher education – and maybe also reconsider just what is meant by that other over-used term ‘democracy’.

  2. AJ
    Posted January 12, 2013 at 11:37 am | Permalink

    That graph might show a correlation, but 1988 and 1976 had the same levels of economic growth but a 10 point difference in vote. I wouldn’t use it as a predictive model.

  3. Adrien
    Posted January 12, 2013 at 11:52 am | Permalink

    Dr Bartels and company think it surprising that Ford got defeated in 1976!!! And according to him Obama’s re-election was typical because of a statistical-historical pattern. I’m sure it’s salient, I suspect Obama would be well-familiar with appearing to understand the underlying mechanics so well. But, um, so…?

    America is a country whose population disagrees on its very constitution. Culturally Romney and Obama represent very different Americas. This is the basis of the apocalyptic rhetoric. Presidents and presidential wannabee’s are just avatars of this dichotomy.

  4. Posted January 12, 2013 at 3:00 pm | Permalink

    [email protected] What am I actually claiming that is so naff? I am simply pointing out that the President’s re-election is not particularly surprising; especially not on economic grounds.

    [email protected] Since not every result fits to the line, clearly other factors matter. But the typical claim about Obama’s re-election is along the lines of “given the state of the US economy then it is surprising Obama was re-elected” when, in fact, the state of the US economy was the biggest single reason to think he would be re-elected.

    [email protected] I don’t think they actually do think it was surprising the Ford was defeated in 1976, you should read the posts. And if the economy was not, in fact, a problem for the President’s re-election, what about it is surprising or otherwise needs special explanation?

  5. Adrien
    Posted January 13, 2013 at 8:41 am | Permalink


    I’ve only read one post; that’s where I got the surprise at Carter’s election in ’76 from. Sorry but the Orient takes most of my attention this week-end. I usually do catch up when I have the time.

    From my reading of the US election (and I paid a deliberately esoteric and minimalist attention to it) the whole thing hung on a knife edge. The polls showed a close one for months leading up to the debates. Commentary on the matter was in agreement that these debates would decide things finally and I suppose they did. Was it, as the chart above is evidence of, especially commonplace? Perhaps according to data which considers only the employment situation and the patterns of incumbency it is. Perhaps the election is ordinary and we can’t see this because of the glamour involved. But, still, the discourse has been hysterical. The very legitimacy of Obama’s presidency has been assaulted since it began. This is all part of the fight of the three disparate groups that make up the Right in America to hamper any attempt to roll back the gains they have obtained since the Reagan presidency (imho).

    In Romney’s case and for advocates of minimal government regulation of the economy everywhere this was a fight against the resurgence of Keynes and the redeployment of New Deal type policies. For flag wavers it’s about Obama’s questionable Americaness and his nominal willingness to roll back the US MIC (we’ll see how that goes). For Bible Bashers it’s his urbane secular liberalism that’s dodgy. In his victory speech Obama spun out the old ‘anyone can make it in America’ schpiel’. The usual dichotomies were deployed: rich or poor, man or woman, black or white, gay or straight.

    Wait! Gay or straight? Them’s fightin’ words in Wichita. True Clinton tried some advocacy for homosexuals. Did he give such a speech on his re-election? I have my doubts. But this is of the essence of Obama’s appeal and at the heart of why people hate him as well. For residents of Urbane America he was the poster boy for Modern Life. For Traditional Values America he was Satan’s secret agent.

    That has a significance that numbers don;t get at.

  6. Adrien
    Posted January 13, 2013 at 9:00 am | Permalink

    I just read Clinton’s re-election speech. He rolls out the old ‘we’ll leave the politics behind and get on with the job God Blessa ‘Merica’; he leaves out the whole liberal diversity riff. Best leave that out. He tried pushing it on the US military before he discovered that 96% of ’em are hardcore fundamentalists. Guess we can’t blame him as he had spent a great deal of time avoiding things military.

    Still the presence of a powerful theocratic ethos in the military? Well how bad could that be? Never mind, only four more years and then it’s someone else’s problem. Tonight! we’re all together:

    Whether you are the party of Thomas Jefferson or the party of Abraham Lincoln, whether you’re an independent or unaffiliated, remember that we all belong to the greatest nation in history.

    Both Obama and Romney were likewise fulla the Greatest Nation In History schpiel. Perhaps I’m being hysterical myself, but it seems to me that the country has serious cultural and structural problems and that, swamped by hubris and wallowing in parliamentary idiocy, the elites in Washington all the way from from the rookie lobbyist thru Congress and Senate to the Oval Office and everyone in it have no real idea that this is so. They assume that their nation will always be the master of the world. The parties of Messrs Lincoln and Jefferson? Dead long time ago.

    In their place is the party of Red Tie Robot and the party of Blue Tie Robot. Both brought to you by the Bland Corporation. Next up the weather.

  7. Posted January 13, 2013 at 2:46 pm | Permalink

    [email protected] First, unemployment is not a major factor in election results. People think it should be, but the evidence is consistent that it is not; this is true also of elections in other developed economics. Income growth is an important factor (and that affects the direction of unemployment up or down), but the level of unemployment, not so much.

    Second, these models are mainly about how the undecideds are going to go (with some side element on whether base voters turn out or not). The further away from an election, the more useless the polls are, since they function more and more like “by-elections” and exaggerate disenchantment with the incumbent.

    Third, the nature of the dynamic of two-Party systems is a tendency towards convergence in vote share. Less so in Presidential elections, since the focus is on two individuals, but it is still there — the above results shows most Presidential elections have a 7%pt or less difference.

    As to how well the US is doing, all things considered, quite well actually. Crime is falling, the debt is manageable (it is still much lower than that the UK began the long period of the Pax Britannica with), the economy is recovering (if anaemically) while the US’s problems are not as bad as Japan’s or the EU’s. They are not grappling terribly impressively with the problems confronting them domestically, but the US is often like that.

    As for their status in the world, that is not going to change soon either. The euro ain’t going to replace the US$ (that ambition now looks utterly pathetic) and China can’t manage to overtake US influence in its own neighbourhood, let alone globally. China’s problem being it worries said neighbours far more than the US does without worrying them enough to be scared of coseying up to the Americans.

    You are right, there is a tendency towards hysterical domestic rhetoric (particularly from the side out of the White House, consider the realities of Bush Derangement Syndrome). It would be good if people generally had a quiet lie down and get over their tendency towards the vapours. I find Obama Derangement Syndrome equally sad. The man is, if anything, a moderate Republican in foreign policy, for example, that’s why he finds it so easy to nominate Republican Defence Secretaries. And, apart from Obamacare (not so different from Romneycare after all), hardly a radical in domestic policy either. Something the American Left has noticed and is not happy about.

    Regarding the US military, they have now swallowed gays openly in the military without much of a bump. Remember, most of their personnel are under 30, and that is the most gays-are-people-too age cohort in the US. While even this Supreme Court may well strike down DOMA, as Congressional overreach. Especially as the Chief Justice seems to be for the long haul and the latest election results are now showing that queer-bashing is losing its electoral cache; a trend unlikely to do anything but accelerate. As Mr Savage cruelly notes, every hearse that turns up to a retirement home is a boon for queer rights.

  8. kvd
    Posted January 13, 2013 at 5:05 pm | Permalink

    [email protected] my ‘naff’ comment was directed to the claim that political analysis is a ‘science’ – not your own comments. I had read your links and was/remain confused. But since you dismiss this I assume you disagree? Let’s look at the graph – just 2008 for instance:

    1) Hard to tell exactly, but it seems, for that election, that the vertical axis “Incumbent Party Popular Vote Margin %” is between -10% and -5% and I’m guessing it is about -7%? The website records the popular votes (in millions) as Obama 69.3 McCain 59.6 Others 2.1 Total 131 million. Question: how does one get a figure of -7% from any combination of the above? In plain English I would suggest McCain (being the incumbent party representative) got 59.6 of a total 131 million – which represents roughly 45.5% of the total vote; i.e. a ‘margin’ of -4.5

    2) Your first link contains the following brief statement: The figure above is derived from the following very simple regression analysis:
    Incumbent Party Margin = 9.93 + 5.48 × Income Growth – 1.76 × Years in Office.

    – which I would parse (i.e. obey the rules of mathematics) as
    Incumbent Party Margin = 9.93 + (5.48 × Income Growth) – (1.76 × Years in Office)
    Honest question: what is the significance of the ‘hard’ numbers 9.93 and 5.48? From the attached commentary I think I understand the 1.76 value, but can’t see the other two.

    3) Again from that link, it seems to be implied that the model makes no pretence to be predictive of future election outcomes – but I would suggest (for your comment) that the implication of any such ‘political science model’ is that one should be comfortable to at least deduce either of the following:
    – the ‘Income Growth Rate’ required to produce an ‘Incumbent Party Margin’ of X%
    – the ‘Incumbent Party Margin’ of (any) X% means the ‘Income Growth Rate’ will have been Y

    Or perhaps I have it entirely wrong (which happens quite often, but if you don’t ask, etc. etc.)? And again, this is simply an enquiring mind – not a personal attack.

  9. kvd
    Posted January 13, 2013 at 5:45 pm | Permalink

    Possible answer to my Q1 above: Obama got 52.9% of the popular vote, and McCain 45.5% as I noted. The difference is 7.4 so I’m guessing this is the ‘margin’ being graphed?

  10. Posted January 13, 2013 at 8:34 pm | Permalink

    [email protected] Yes, it is the margin in %pts between the Republican and Democrat candidates.

    [email protected] Yes, the ‘science’ title is disputable but I am only concerned with the models themselves.

    Regression analysis such as this has a ceteribus paribus element to it. If the model “gets it right” then the other factors “cancelled out”. If not, then factors not covered by the model had an effect.

    5.48 is the weighting given to income growth. 9.93 is the “intercept”. It may or may not have significance. In this case, it might be the level of irreducible support you get for showing up as the incumbent Party candidate. Though it’s late and I am not very mathematical, so perhaps not.

  11. Movius
    Posted January 14, 2013 at 12:56 am | Permalink

    I find Dan Carlin’s explanation for the 2012 US presidential election result compelling, if a little simple.

    That is: When it comes to the economy, Americans are well aware that Obama is steering the good ship Titanic. The Republicans erred by picking a candidate who was seen as steering the iceberg.

  12. Adrien
    Posted January 15, 2013 at 10:24 am | Permalink

    Lorenzo – First, unemployment is not a major factor in election results.

    Ever? The seven figure sum out of a job in Germany at the beginning of 1930 had no effect on the vote in March? Sorry, being facetious.

    So it’s not unemployment figures its expectations of future security as expressed by how much you can spend right now. And, lucky ol’ ‘Bama had the money-to-spend graph spike up, so he won. This is even tho’ according to that chart real income wasn’t back to where it had been at the beginning of the year.

    What struck me about the Monkey’s Cage’s Trend and Perceptions chart is how closely peoples’ general perception corresponds to how much I have right now. I like the brave spike of optimism in mid-February.

    Interesting also how the undecideds become economically pessimistic earlier than those whose minds are made up. They are also more cautious about recovering their optimism as income ascends.

    That’s assuming that a measurement of perception can truly be reliable. I have doubts. But economic perception would probably be reasonably easy to get a response to it. The model suggests a constituency of goldfish whose vote depends on how much Bud they can buy.

    The man is, if anything, a moderate Republican in foreign policy, for example, that’s why he finds it so easy to nominate Republican Defence Secretaries.

    Obama models himself on GOP presidents past, particularly Lincoln, (which is a worry.) Bringing your enemies into cabinet is a Lincoln riff. I reckon he’s really more like Kennedy. The Democrats needed a new pretty poster boy for Liberalism. And Hollywood provided.

    Remember, most of their personnel are under 30

    I’m not referring to most of their personnel. The US military is not a democracy.

  13. Adrien
    Posted January 15, 2013 at 11:39 am | Permalink

    Apologies for following ugly bit of non-English: But economic perception would probably be reasonably easy to get a response to it.

    What I mean was that economic perception’s probably more reliably measured than other kinds because you can count the money in your pocket, etc. I was in a rush.

    Viz: the assumption of continued US prestige, I don’t mean to suggest that, as a state, the United States is due for some sudden long-sliding slump. But I do believe it has serious problems, these may not destroy the nation but they have potential to transform it into something else.

    Apart from the lurking anti-secular sentiments and the destructive decadence of political machination and special-interests distorting fair government there are also the inevitable geo-political challenges to American supremacy. It’s unwise to assume that the US will meet these latter challenges and be always victorious. There are good reasons in fact to suspect the opposite.

  14. Mel
    Posted January 15, 2013 at 11:41 pm | Permalink

    Has a date for Kick A Ginger Day been announced for 2013? I have new boots and I’ve been in training for six months and I’m really looking forward to this years punting my way through a sackful of gingers.

    Other than that I have nothing to report ….

  15. Mel
    Posted January 16, 2013 at 8:50 pm | Permalink

    lol. American rape crisis feminist Sara Ylen may end up being charged over one of her numerous false rape allegations.

    Also note this: “Separately, Michigan State Police Detective Sgt. Brian Ferguson is conducting a fraud investigation involving Ylen. Details about that case have not been released.”

    I wonder if the fraud case involves cancer? Time will tell I guess.

  16. Posted January 16, 2013 at 10:18 pm | Permalink

    [email protected] I am not sure comparing Obama to JFK is exactly a rebuttal to my suggestion his foreign policy is moderate Republican.

  17. kvd
    Posted January 17, 2013 at 3:38 pm | Permalink

    Since this is denoted as a chit chat post I’d like to copy you with a comment I made to the ACCF just now. I understand that half the world’s population posesses a cervix, and I understand we should rightly honour our researchers – but nevertheless:

    This is a difficult memo to write to you because as a layman whose wife died from a brain tumour I have nothing but admiration for our researchers in all cancer-related fields. And most particularly for Professor Ian Fraser for his work in your field.

    Last night I received a ‘cold call’ “on behalf of the Australian Cervical Cancer Foundation” soliciting funds. Not unexpected; have received the same, and donated, a number of times. But the emailed receipt for my (minor) donation has some very small, grey-coloured detail at the bottom which makes it plain that only 40% of my donation was actually going to the ACCF – with the rest to some fundraiser company.

    Annoyed by that, I went to the ACCF website (The “Australian” CCF) only to find that it seems the majority of funds raised by your organisation are in fact directed to overseas locations, with something termed an “education campaign” being the only local (i.e. Australian) effort worthy of reporting.

    I feel this is a misuse of the word “Australian”. Further I feel that receiving a call “on behalf of” the ACCF from a company which stands to receive 60% of my donation is misleading.

    I will be happy in future to continue direct support of your organisation, in direct support of Australian-directed initiatives (indigenous issues, for instance) but wanted to register my complete disillusion with your present fundraising method – and advise that you have otherwise lost a supporter

    I guess I should emhpasise that the receipt for my donation, and the ACCF website are both perfectly ‘up front’ about what is happening. My objection is simply that none of this is actually ‘up front’ to somebody receiving a cold call at 8 o’clock at night for an otherwise very worthy-seeming cause.

  18. Adrien
    Posted January 19, 2013 at 2:19 pm | Permalink

    I am not sure comparing Obama to JFK is exactly a rebuttal to my suggestion his foreign policy is moderate Republican.

    Lorenzo my comparison has nothing to do with policy and everything to do with imagery.

Post a Comment

Your email is never published nor shared. Required fields are marked *