Comment: Why Obama is going to lose

If you believe the polling data coming from the US, you are probably convinced that Obama has at least 80% chance of winning. Prepare to be surprised.

I have lived in the USA for 13 years and have followed closely the American politics for two decades. I find that here in the UK there is a lot of misconception and misinformation about the US' internal politics.

A recent experience brought this point home, when, a few weeks ago, at the Conservative party conference in Birmingham, the vast majority of the delegates attending the American election special fringe event, jumped to their feet to cheer at the thought of Obama's re-election.

How could so many Conservatives love a politician who is the American equivalent of our own much-despised Ed Miliband?

To me, this big disconnect between the perception and the reality indicates the lack of objective information available to the British news readers, at least as far as the topic of the US internal politics is concerned. How much of it is intentional and how much of it is simple oversight, hard to quantify.

It appears that British media (even the right-wing) and British public are not scrutinising enough the information sourced from the US mainstream media, which is assumed to be reasonably objective and unbiased. Reporting, based on biased input, will inevitably produce biased output.

As a result, the majority of even British Conservatives expect and hope that Obama will win.

Most people may not be aware that the whole of US mainstream media, with the exception of Fox News, are heavily left-leaning, dedicated to promoting the leftist agenda and are heavily invested in Obama's re-election.

US mainstream media 'involvement' in the electoral outcomes is huge and well documented. Several major media personalities had publicly estimated that media's support gives Democrats about 15% electoral advantage. The analysis of the data of 18 past US elections (since 1940) shows that only one Democrat (Johnson in 1964) won the presidency with a margin of over 15%. 

This demonstrates that without the constant support of the mainstream media, the US political left would be a permanent minority party in the US political scene.

Mainstream media's support for the democratic candidates materialises in several ways:

• By over-emphasizing every faux pas and mistake of the Republican candidate, even if it's a minor one. If they cannot find any 'dirt' on a candidate, focus on minor issues, take them out of context and overblow them.

• By highlighting every possible success of the Democrat candidate and ignoring or underplaying all damaging stories and scandals involving them. If not reporting a negative story cannot be avoided, then report them as 'this is not a big deal' or 'this is not really as bad is it might sound'.

Here are a few memorable examples:

• Republican Richard Nixon was forced to resign because he lied to the investigators of the Watergate break-in. But when Monika Lewinsky sex scandal broke out in the late 1990's and Bill Clinton was caught lying about it to the grand jury, numerous stories appeared in the US mainstream media about how lying is actually not always a bad thing… Lies can spare the feelings of the people close to us… That's a good thing.

• When Bill Clinton became president in 1993, he fired all attorney generals appointed by the prior Republican presidents. Nobody in the media had a problem with that, and rightfully so, as appointing and firing attorney generals is president's prerogative and most presidents do it. Eight years later, after a contentious post-election fight in 2000, the new president, G.W.Bush, in the somewhat misguided spirit of reconciliation, decided to keep almost all of Clinton's attorney generals except two or three. The press erupted into a scandal as if Bush had just arrested congress and declared martial law, thus ending the 200+ years' tradition of democracy. It was the all-consuming front-page story for months. Can I say 'double-standard'?

• On a similar note, in the mid 2000's, when the US economy was booming and the unemployment rate was 5.4% (which is considered statistical full employment), US mainstream media went on a multi-year campaign of trying to convince everybody that the country was in a recession and many unemployed people were living miserable lives. When, during the Obama’s presidency, the unemployment shot up to ten per cent (double what it was just a few years earlier), numerous heart-warming stories appeared about how wonderful it was to be unemployed, how those, who had recently lost their job, could now relax, being no longer stressed by the rat race, how they now had time to pursue their interests, hobbies, rekindle their relationships.

You may wonder at this point, how relevant this is to us? I think important lessons and insights can be learned from all this that could make a difference during the expected uphill battle for the right in the UK in 2015.


If one were to believe the accuracy of the polling data coming from the US, one would conclude that Obama's victory is almost inevitable. Polling data, however, are not accurate. It's biased and carefully engineered to give a very distorted sense of reality.

The majority of mainstream media in the US (with the exception of Fox News) are so emotionally invested to elect a democrat (Obama), that they use the polls not to report, but to shape the public opinion by skewing the polling results in favour of the democratic candidate.

Reporting the poor polling performance of the Republican candidate reduces Republican voter enthusiasm. This has two consequences:

1. It leads to a reduction in campaign donations ('what’s the point in giving him money if he’s going to lose anyway?'), resulting in the campaign being able to buy fewer political ads.

2. Reduced voter enthusiasm causes lower voter turnout.

How can numbers lie?

The intentional distortion of the polling results is achieved by skewing the statistical sampling.

Creating valid statistical sample is a complex art and science. It is easy to fabricate results by creating samples that over-represent certain population (oversampling): since most registered Republicans will vote Republican, and most registered Democrats will vote Democrat, one simple way to skew the polling results to the left is by polling a greater percentage of Democrats than are likely to vote in the upcoming elections.

The two most common ways of manipulating the statistical sample are:

• Polling registered voters instead of likely voters (statistically, registered Democrat voters are less likely to vote)

• Base the expected voters’ turnout figures on the last presidential elections that had an unprecedentedly high Democratic voter turnout.

In 2008, in addition to suffering the negative legacy of Bush's administration, John McCain's campaign was anaemic, uninspiring, and at times seemed as if it was designed to lose nobly to the nation's first minority presidential nominee. Enthusiasm amongst the Republican voters was at all-time low, while Obama was inspiring unprecedented enthusiasm even among people who were politically apathetic and usually didn't vote.

In 2012, following the much-maligned G.W. Bush, Obama started his presidency with a 70% approval.

He was perceived as 'the ONE', a much awaited messiah who would heal all wounds and political divisions, eliminate racism, re-unite the country, fix the economy, balance the budget, cool down the climate, lower the ocean levels, solve all international conflicts by using his magical persuasive powers, and make the USA loved and adored all over the world again. The sentiments and the optimism were so exuberant and pervasive, that only nine months after his inauguration, he was awarded the Nobel Peace Prize for his "vision" and for "giving the world's people hope for a better future".

The last election was all about words that inspired the nation.

This election, on the other hand, is more about un-kept promises and disappointing results: unemployment rate, national debt, GDP, etc.

During his four years in office, Obama hasn't solved any problem, and he's made some of them worse:

• Obama's key election promise was to get the economy working. He failed on this task. Economy has not recovered.

• Debt has exploded by $6 trillion. Instead of halving the deficit, as he had confidently promised during the campaign, he doubled it by embarking into an unrestrained spending campaign that would have made both Gordon Brown and Ed Miliband tear up with pride.

• Unemployment is higher than when Obama was elected four years ago.

• Petrol prices have doubled.

• Average family income is down by almost ten per cent.

• The country is politically polarised more than when he came to power.

So, given all this, what's the best way to estimate the voter turnout in 2012?

A much more accurate voter turnout picture would be based on the 2010 mid-term elections, with a lot of motivated anger on the Republican side and quite a bit of disillusionment with Obama on the Democrat side.

As an example, let's consider Ohio, one of the key swing states. Obama won Ohio in 2008 by eight per cent margin, but in the 2010 congressional elections Republicans won Ohio by one per cent. That's a n ine-point swing that mainstream media pollsters are not taking into account when composing their polling statistical samples.

The polling results, which have been reported up until recently, had been consistently oversampling Democrats by as much as seven to eleven points. And if Obama in those polls is only up by three to four points, then in reality Obama is down by three to eight points! Suddenly the situation turns from very likely win for Obama to almost certain win for Romney.

Now enthusiasm for Romney is 30% higher than it was for McCain at the same point in time during the 2008 campaign, while Democrats' enthusiasm for Obama is down by eight per cent. That's a huge shift that most polls don't account for.

What I find amusing is how the polls start moving to get in line with actual results when the actual election date draws closer. There is a reason for that. While it is easy to manipulate the polling results when nobody can prove that you are wrong, it's much harder to dispute the definitive election results.

So a time comes when polling agencies' focus shifts from helping their candidate to protecting their post-election credibility. If the pollsters' predictions were found to be way off the actual elections results, their credibility (hence job security and ability to influence future events) would be heavily jeopardised. So, when the election time gets closer, typically within a few weeks before the date, all major pollsters start moving their polling results closer to reality, so that, when the votes are counted, their predictions will look close enough not to attract ridicule.

By then, most of the damage to the R epublican candidate will be already done, a lot of it irreversibly: while voter turnout may have partially recovered, the campaign fundraising and political ads side will have been harmed permanently.

In 2012, we started seeing this polling shift in early October, about a month before the elections. This reversal of fortune was universally attributed to Obama's poor performance during the first presidential candidate debate. In truth, though, that's not what started it.

The first debate only added to the momentum of the process that had been in play for quite a while already, and it provided the mainstream media with an optimally timed explanation for starting to move their polls closer to a more likely scenario.

Without a doubt, the first presidential candidate debate did play a role, but it's only one of two major factors responsible for what seemed like a change of heart of the electorate.

In normal elections, all polling organisations would be migrating towards the more realistic results by now, but this time there is a split: while a good number of pollsters did do that, quite a few of them did not, and continue to use the 2008 turnout sample. Why is that? Because, in the eyes of many in the mainstream media, Obama is still such a unique transformational figure, that he must be protected at all costs, more so than any Democratic presidents that came before him. Some of them are so desperate for Obama to win, they are setting aside their instincts of self-preservation and are willing to go down with the ship if that gives hope of saving the political life of captain Obama. They are willing to put their reputations and their careers on line for him.

There is an additional explanation for this, though: since it's very likely that, through internal polls, Obama's campaign are fully aware that they are in deep trouble, they are preparing to file lawsuits challenging the outcomes in battleground states, alleging Republican voting fraud as a reason for the discrepancy between the polls and the votes.

Obama's election prospects

Generally speaking, as far as Obama's chances of re-election are concerned, things for him are not looking good: while I can't find a single credible (i.e. unbiased) metrics that would give Obama a clear advantage, a number of predictor models, which had worked reliably in the past, predict victory for Romney. Here is just a partial list of those:

– All major likely voter polls, once compensated for Democrat voter oversampling, have been giving Romney clear advantage for months. In the last few weeks, quite few of them, even with Democrat oversampling, show Romney as being ahead.

– Two Colorado University political science professors (Kenneth Bickers of CU-Boulder and Michael Berry of CU Denver) have developed a model that predicts presidential elections results based purely on economic data. They've been using their model since 1980 and they haven't been wrong thus far. According to their most recent analysis, Romney should get 330 electoral votes vs Obama’s 208.  That's a further 5 votes shift towards Romney from their 22 August prediction (Romney 325, Obama 213).

– Another model, that has worked reliably in the past, bases its prediction on the presidential approval numbers in the swing states. According to this, Obama’s approval numbers are slightly worse than Jimmy Carter’s in 1980 (when Carter lost in a landslide 489-49).

– 70 million people tuned in to watch the first presidential candidate debate. That's a larger than usual number, indicating the dissatisfaction with the status quo. Another poll indicates that 63% of Americans want change. That is always bad news for the incumbent.

– In 2008, through the unprecedented electorate excitement and media's celebration, Obama won by 7.3%. Respectable, but not huge. This time Republican voters’ enthusiasm for Romney is 30% higher than in 2008, while Democrats' enthusiasm is down by 8%. If one were to adjust the nationwide polling data for this factor alone, Romney comes out ahead by possibly as much as ten per cent. Ronald Reagan beat Jimmy Carter in 1980 by 9.75%.

– Among likely voters, Romney leads Obama by 10 points (53%-43%) among men and nine points (52%-43%) among independents. Obama used to have advantage among women, but even that one has evaporated.

– Another indicator of the changing voter enthusiasm is the sizes of rally crowds the two candidates are attracting. Obama events attendance is 1/10th of those in 2008. Romney’s have been growing.

– Finally, there is a mistaken perception that incumbency dramatically increases one's chances of winning. The reality is quite opposite, especially for Democrats: in the last 68 years only one Democrat incumbent has been re-elected (Bill Clinton – and his survival strategy was in many ways was opposite of what Obama has done). Republicans won four re-elections.

This has been one of the most animated campaigns in recent history, and its outcome may still appear to be uncertain. Based on my research, however, I see no reason for optimism in Obama's camp. Just like Jimmy Carter before him, he's likely to end up a one-term president.

Is this a good or a bad thing for America and the world? That is another story. Stay tuned!

Gintas Vilkelis is a Lithuanian physicist who has lived in the US for over a decade and now resides in Buckinghamshire where he is a member of Stoke Poges Conservative Committee and the Beaconsfield CCA executive council. He has worked at Lawrence Berkeley Lab, Brookhaven National Lab as well as at the Geneva-based European Centre for Nuclear Research.

The opinions in's Comment and Analysis section are those of the author and are no reflection of the views of the website or its owners.