Tag Archives: beer test

Why I Might Buy Maureen Dowd A Miller Lite

Dear Maureen,

(May I call you that?)  I understand that you probably don’t read my blog. You are a busy person, after all, with your column and all that goes with being a celebrity journalist (may I call you that?).   Your time is undoubtedly precious.  But do you read your own paper – the one that publishes your column?  Not the whole paper,  mind you, but maybe just the op-ed page.  You know, that section that publishes opinion pieces (like yours)?

I ask because a reader sent me a link to your August 11 New York Times column, titled “Likability Index”. In it, you discuss whether President Obama and Mitt Romney are truly likable people.   You claim that Obama is a lot more likable, at least when viewed from a distance, than is Mitt. This likability gap is important because,  you suggest, it will determine who wins the presidential election: “The big difference, the one that will probably decide this presidential race, is this: Barack Obama is able to convey an impression of likability to voters. Given how private he is, an enigma even to some who are close to him, it’s an incredible performance.”

Yes, Obama “can be thin-skinned and insecure at times, but he radiates self-sufficiency, such a clean, simple aesthetic that he could have been designed by Steve Jobs — Siri without the warmth.”  As for poor Mitt, well, here are your words: “Romney started out off-putting and now makes Willy Loman look like prom king. Obama is introverted and graceful; Romney is introverted and awkward.”  Ouch!

I realize you have some data to back up your claim of a likability gap; as you note, there’s extensive polling evidence indicating that Obama has “better numbers on honesty, trust and empathy” in addition to enjoying a sizable advantage in likability. Your conclusion?  “Once a candidate gains the advantage in ‘Who do you want to have a beer with?’ — even if he doesn’t drink beer — it’s very hard to reverse.”

Oh no! Not the beer test!  I thought I debunked the myth of the beer test in this post, when – drawing in part on research by three prominent political scientists – I argued that there was no strong relationship between the candidates’ relative likability and the outcome of the presidential election during the period 1952-2000.  I concluded: “The point here is one I’ve made before: presidential elections are driven by fundamentals – national conditions and candidates’ issue positions – far more than they are by the candidates’ personal qualities. Indeed, I don’t know of a single reputable presidential election forecast model that incorporates likability ratings.”  Clearly, however, you aren’t a regular reader of my blog.  But surely you read your own paper?   If so, you might have run across this piece on your paper’s very own opinion page written by one of those prominent political scientists, in which he argues “that a candidate’s likability is a relatively minor factor in deciding modern presidential elections.”

Once again, in case you missed it when it was published in the Times, here is the summary data in graphical form:

It shows that, historically, the candidate rated higher on the personal dimension did not always win the election.  Indeed, in 1996, Bill Clinton – the most negatively rated candidate of all during this period – trounced World War II veteran Bob Dole, while in 1980 voters soundly rejected Jimmy Carter despite his sterling personal qualities.

Of course, 2012 might be different.  Maybe, in a very close election, enough voters’ choices will turn on whether they prefer to have a beer with Mitt (non-alcoholic, of course) or the President to swing the election one way or another.  If that happens, here’s my promise to you Mo (may I call you that?):  I’m buying the next round.

Do you drink Miller Lite?



(Yes, you can call me that.)

Romney or Obama: Who Wins the Beer Test?

The other day I was watching my go-to source for informed political analysis, ABC’s The View, where the topic du jour was whether one would prefer to have a beer with Mitt Romney or President Obama.  This “beer test” question, of course, is supposed to measure a candidate’s likability factor.  How well does he empathize with the common man?  I first heard the question used to explain why Al Gore barely won the popular vote, and lost the Electoral College, to George Bush in 2000, despite running as the incumbent party standard bearer during a time of peace and prosperity.  Gore may have been smarter, the pundits suggested, but the thought of listening to him drone on about the social security lockbox was enough to swing more than a few votes to Bush. Simply put, more people preferred to down a pint with George than with Al.

Pundits believe strongly in the likability factor, and they often trot it out to explain why a candidate did unexpectedly well, or not so well.  Consider National Review’s Jonah Goldberg explanation for Michael Dukakis’ defeat by George H. W. Bush in 1988: “At least one of the myriad reasons for … Dukakis’s loss in 1988 was that he seemed like the kind of guy for whom the best time of his life was when as VP of his high-school chess club, he ascended to the top spot when the president got mono.”  Of course, almost everyone cites the beer test as an explanation for why Bush the Younger bested John Kerry in 2004. This USAToday columnist even cited polling data to make the case for Bush as a better beer companion that year: “President Bush, despite his many problems, strikes most of the American people as a pretty nice guy — the kind of guy they would feel comfortable with if he showed up at their front door. The more standoffish Kerry projects little warmth. A recent Zogby/Williams Identity Poll reflected that. It found that 57% of undecided voters would rather have a beer with Bush than Kerry. (In Bush’s case, it would be a nonalcoholic beer.).”

It’s intuitively appealing to believe that the candidate who is viewed as more likable is likely to do better in the presidential election.  And, all other things being equal, it is probably better to be viewed as likable rather than unlikable. But when we look at the factors that drive election results, likability simple hasn’t proved very predictive, as Stanford political scientist Morris Fiorina pointed out last week in this New York Times editorial. Along with Sam Abrams and Jeremy Pope, Fiorina compared the public’s evaluation of presidential candidates’  personal qualities (separate from policy stances or experience) with election results in the period 1952-2000. Their conclusion? As Fiorina wrote in his op ed piece: “Over all, in the 13 elections between 1952 and 2000, Republican candidates won four of the six in which they had higher personal ratings than the Democrats, while Democratic candidates lost four of the seven elections in which they had higher ratings than the Republicans. Not much evidence of a big likability effect here.”  This table shows the relative standing on the personal dimension of the major party candidates during the 13 elections from 1952 through 2000:

As Fiorina suggests, there’s no immediate pattern linking higher ratings on personal qualities to victory in the election. Moreover, Fiorina notes that in 1980, Jimmy Carter was rated the highest on personal qualities among all Democratic nominees in this period, and yet he was soundly trounced by Ronald Reagan. In 1996, on the other hand, Bill Clinton was the lowest rated candidate – either Democrat or Republican – to run in the 13 elections Fiorina studied, and yet he easily bested World War II veteran Bob Dole.

The point here is one I’ve made before: presidential elections are driven by fundamentals – national conditions and candidates’ issue positions – far more than they are by the candidates’ personal qualities. Indeed, I don’t know of a single reputable presidential election forecast model that incorporates likability ratings.  Knowing which candidate the public prefers to have a beer with isn’t going to be very helpful in predicting who is going to win the election. Alas, this is a point that will undoubtedly get lost as pundits like those on The View begin applying the beer test to judge the likability of Mitt Romney or Barack Obama.  For what it is worth (and my point is that it is not worth much relative to other factors), Obama has the likability edge over Romney, at least based on recent Gallup poll.

His edge extends to independents, but not Republicans.

And yet, at the same time this poll was in the field, Gallup’s daily tracking polls had Romney and Obama in essentially a dead heat.  The reason why, I suspect, has something to do with the split among independents on questions 4 – agrees with you on the issues – and 5 – who can manage government more efficiently.  The answers to those questions, history indicates, will tell us more about the election outcome than will the likability gap.

For what it is worth, I’m not sure I want a beer with either candidate. Mitt, who doesn’t drink, would just make me feel guilty as the night dragged on and he sipped on his non-alcoholic beverage.  And Obama might be sitting at my table, but I’m guessing he wouldn’t be talking to me so much as at me.  With either one, then, it would be a long night, and not in an enjoyable way.

All in all, I’d rather drink alone.