All three of the presidential candidates now leading the Democratic primary field are 70 years of age or older.
Thirteen of the remaining candidates are 60 or under.
These facts have been much written about, but do they matter?
You might reasonably turn to the polls for an answer. After all, campaigns analyze polls to determine which candidate attributes to focus on, while pundits scour the results to predict which candidate will triumph.
My purpose here is less to explore the impact of candidates’ ages, and more to examine how we reach conclusions about these matters from polls.
Such “analysis” is not nearly as straightforward as it appears.
A July YouGov poll report concluded, “Democrats … are clear about one ideal trait: relative youth.”
More than three-quarters said their ideal president would be under 60, while fewer than 4 percent preferred a candidate 70 or older.
A Pew poll produced similar, but not identical results, with 47 percent of Democrats saying it’s best for a president to be in their 50s, with just 3 percent opting for a candidate in their 70s.
These findings would appear to spell big problems for former Vice President Joe Biden and Sens. Bernie Sanders (I-Vt.) and Elizabeth Warren (D-Mass.), which might seem strange, since they occupy the top three places in almost every poll of the primary race.
Are the polls wrong? How is it that voters overwhelmingly want a candidate under 60, and have a slew of such choices, but opt for the candidates in their 70s, seemingly their least favored alternative?
Of course, the poll questions examined so far ask for preferences but, as is too often the case with polls, don’t ferret out the importance voters attach to those preferences.
One could prefer a president in their 50s to one in their 70s, but not care much at all.
To find out how much people care, pollsters too often simply ask people how important a factor, like age, might be to them.
IPSOS asked just such a question and found more than half of Democrats claiming they’d be less likely to support a candidate over 70 years old.
However, as I’ve argued here before, and will again, there is absolutely no evidence that answers to these kinds of questions tell us anything at all about what is truly important.
Several publicly available studies employ conjoint analysis, a technique with far greater reliability, to assess actual importance to voting decisions. In these experiments, voters are presented with a series of match-ups between hypothetical primary candidates, randomly varying several traits for each, including age.
CBS News and YouGov’s non-technical analysis concluded that “voters strongly preferred younger and middle-aged candidates to candidates in their 70s.”
Across all the CBS/YouGov conjoints, voters were about 20 points more likely to select candidates in their 40s and 50s than those in their 70s, suggesting that, at least compared to other demographic factors, candidate age is important.
We’re still left wondering why the septuagenarians are dominating the field.
We don’t know how much age matters compared to those factors.
More broadly, moving from the generic to the specific is not always clear cut. In our example, while 70 may be too old, not all 70-year olds are too old.
That’s exactly what a new CBS YouGov poll on candidate age tells us.
In response to questions specific to each candidate, just 37 percent said Sanders was “too old” to be president, 31 percent said Biden was and only 5 percent believed Warren is too old to be president.
Polls seem to tell us that age is important. But failure to ask the right questions, in the right ways, can lead analysts and strategists astray.
Mellman is president of The Mellman Group and has helped elect 30 U.S. senators, 12 governors and dozens of House members. Mellman served as pollster to Senate Democratic leaders for over 20 years and as president of the American Association of Political Consultants.