Barack Obama's victory over John McCain came in large part from his near-unprecedented support from America's young people— 66 percent of 18- to 29-year-olds and 69 percent of first-time voters cast their ballots for Obama in 2008, compared with 53 percent of the population overall. Despite getting trounced in the midterm elections last month, the Democrats have maintained their edge with young voters, holding on to 58 percent of the under-30 set, according to exit polls.
A newly released study suggests that the echoes of Obama's youthful appeal could continue to reverberate through 2012 and beyond, simply because voters who register for one party keep casting ballots for that party for years after. The first-time voters who were caught up in Obamamania when they registered are likely to stay Democrats well into the future.
In the new study, economists Ethan Kaplan and Sharun Mukand compared the political allegiances of Californians who turned 18 just before and just after the 9/11 attacks, which caused a national shift to the right. They found that voters with birthdays in September were more likely to register as Republicans than voters with birthdays in August. These voters then continued to register as Republicans at higher rates in 2006 and 2008. Once a registered Republican, always a Republican, it would seem.
On the face of it, there's no obvious reason why the mere act of registering for a party need have a lasting influence on one's political sympathies. It's a straightforward process to change your party affiliation, and it's a virtual necessity for anyone moving residences. Typically voters who move need to re-register in order to vote near their new address, at which point they're presented with the option of changing their registered political allegiance.
Yet there are reasons to expect party ties to be sticky. A registered Republican will find himself on all sorts of mailing lists that expose him to an onslaught of propaganda explaining the evils of taxation and the glories of laissez-faire capitalism; registered Democrats are similarly treated to propaganda about affordable health care and regulating Wall Street. The net effect is a hardening of party affiliations on both sides.
Once you become part of the Republican club, your Republican identity becomes part of who you are. There exists an innate human desire to fit in, to have what psychologists call a social identity. It's a powerful enough force that experimenters have been able to manipulate the way total strangers treat one another simply by assigning them to different-colored "teams." Those randomly assigned to be Red sympathize more with other Reds and are more antagonistic toward Blues. Similarly, registered Republicans interpret the news and see the world in a way that's sympathetic to Republican views and hostile to Democratic ones, simply because their Republican registration becomes part of their identity.
Still, there is a remarkable connection between the ballot cast by a voter today and the politics of his youth: Voters who turned 18 during the Kennedy era are more likely to vote for Democrats than those just a few years their senior, who came of age in the (Republican) Eisenhower years.
There are many possible explanations for these patterns. Voters obviously don't register as Republicans by random chance—they do so at least in part because they have Republican views to begin with. The fact that those views often last a lifetime may have nothing to do with a feeling of kinship with other Republicans but rather with the fact that these voters possess firmly held Republican beliefs. Such party loyalists might also have lasting family circumstances that lead them to favor Republican policies. For example, many rich kids register as Republicans at 18 and continue to register as Republicans when they're rich 40-year-olds. And growing up at different times—the counterculture '60s vs. the conformity of the 1950s—presumably gives the youth of these eras very different formative experiences that shape their politics forever.
To strip away these complications, Kaplan and Mukand compare the registrations of California voters born between Aug. 16 and Sept. 15, 1983. Because of their slightly earlier birthdates, the August babies were more likely to register before the 9/11 attacks than the September babies. (Registering to vote for the first time may also be the kind of thing you're more likely to do in a timely fashion if your birthday is before the beginning of the school year. In 2001, Aug. 31 was the Friday before Labor Day.) By focusing just on the effect of slightly different birthdays rather than comparing those who registered after 9/11 to those before, the authors avoid the problem that different sorts of people—perhaps very Republican ones—may have been prompted to register by the terrorist attacks.
The slightly younger September cohort thus only got around to registering as the Bush administration's war on terror was underway. The homeland security message resonated with a public wary of further attacks, gave a boost to the approval ratings of George W. Bush, and presumably influenced the registration decisions of young voters.