Do New York City voters know their own minds?

New Yorkers like term limits. City residents have voted twice to support a two-term limit for local officials, first in 1993 and again in 1996. And twelve years later, according to a poll conducted just last month by the Quinnipiac University Polling Institute, they still hold that view.

But in that same poll, a majority also said they were in favor of waiving term limits so that they could vote for Mayor Mike Bloomberg again in 2009. If voters thought that Bloomberg should get a third term, why didn’t they take the next logical step – that term limits are a bad idea? Is a sizeable portion of New York schizophrenic? Or are voters living out F. Scott’s Fitzgerald’s famous dictum, that “the test of a first-rate intelligence is the ability to hold two opposed ideas in the mind at the same time”? What’s going on here?


For political scientists, the answer is: Nothing special. People who study public opinion are accustomed to polling results that are internally inconsistent. “You find this thing a lot,” said Robert Shapiro, a professor of political science at Columbia University, especially when voters’ “actual opinion about the issue is complicated – there may be some ambivalence.” Because polls generally sort opinions into “either/or” categories, that complexity may show up as inconsistency. The effect can be pronounced when voters themselves are unaware that their views are in contradiction, and when a poll’s design does not bring the contrast to their attention. (The Quinnipiac poll proceeded from an abstract question about term limits to a specific question about Bloomberg; had the order been reversed, some respondents who favored a third Bloomberg term might have rethought their opinion about term limits in general.)

But that explanation only says that people sometimes hold contradictory views; it doesn’t tell us why that might be so. An answer to the underlying question may be found in another discipline, the related fields of cognitive science and behavioral economics. As a group, researchers in these fields are more likely to work in business schools than political science departments, and they have spent more time understanding consumers’ choices about orange juice and skin-care lotion than voters’ preference about candidates. But their understanding of how human decision-making does – and doesn’t – work can help to puzzle out the polling paradox.

Behavioral economics helps to explain why people don’t always act in rational ways – why governments and businesses are more inclined to keep investing in a doomed enterprise if they’ve already sunk a lot of money into it, or why you’re more likely to use your credit card at the gas station if the lower price for paying with cash is referred to as a “cash discount” than if it is described as a “credit surcharge.” One of the field’s key explanatory tools is “loss aversion,” the human tendency to assign a higher value to the costs of an exchange than to its potential gains. A corollary of loss aversion is “status quo bias,” the preference to remain at the status quo rather than choose an alternative.

Status quo bias explains why designating something as the default option – as we do when requiring people to buy car insurance, or encouraging them to select a retirement plan – makes it more likely that people will settle on that choice. It also helps us understand why incumbents are so hard to beat – and why incumbents in a crowded field are even harder to topple. As the scholars Jack Knetsch, Richard Thaler and Daniel Kahneman note in a review of the relevant academic literature, “the advantage of the status quo increases with the number of alternatives.” That explains why, faced with a bewildering array of whole-grain, fat-free, grass-fed, Omega-3 options in the supermarket aisle, you might reach for the box you remember your mom bringing home 25 years ago. And it’s one reason Bloomberg did so well in the Quinnipiac poll, which asked voters to choose between him and six lesser-known potential challengers, each of whom polled within a few points of each other. (“Don’t know” received just as many responses as the voters’ second choice, Brooklyn Borough President Marty Markowitz.)

But if status quo bias tips the balance in favor of the default option, what happens when two “defaults” – in this case, a popular two-term mayor and a popular two-term limit – come into conflict? Logically, one should be sacrificed. But as the polling results show, they sometimes sit uneasily beside each other.

“There is a status quo bias operating in both the case of the incumbent … and in the case of not wanting to lose the two-term restriction rule,” said Knetsch, now a professor emeritus at Simon Fraser University. “In both, people view the change as a loss of what they have for the lesser-valued gain.

“So when asked if they want to change the rule,” he continued, “they weigh the loss more and tend to say no; and when asked about allowing a third term for Bloomberg, they say yes because to say no is to incur a (more heavily-weighted) loss of present mayor.”

Ironically, public support for term limits can be understood as the voters’ recognition that they are prone to keep politicians around past their expiration date – in the words of Rice University political scientist Robert Stein, they are “a lazy voter’s approach” to protect voters from themselves. It’s “a type of self-control problem,” said Thaler, who teaches at the University of Chicago’s Graduate School of Business. “Voters like term limits in the abstract precisely because they realize that incumbents have a big advantage… So it is not surprising that voters prefer Bloomberg to a bunch of unknowns – that is the preference they were trying to protect themselves from!”

The case of Ronald Lauder, the billionaire cosmetics heir who bankrolled two term-limits campaigns during the 1990s, makes the point. After initially opposing an exception for Bloomberg, Lauder relented – on the condition that he be appointed to a commission that will work to restore term limits through another referendum in 2010, effectively granting a “one-time-only” waiver to the mayor (and a fortunate group of legislators who would otherwise be forced out of office this year). It’s the equivalent of an addict who sets up a complex system to keep himself from his drug of choice. But when a crisis hits, he’s ready to indulge one last time before swearing off his addiction again – this time, he promises, for good.


There is another irony to this story. While voters’ views on term limits are a jumble, the effect of those views may be, at some level, pragmatic. “The most realistic effect of term limits,” the law professors Elizabeth Garrett and Rick Pildes wrote recently, “may be to raise the bar for officials seeking to serve beyond a pre-set number of terms: they must be well-regarded enough to overcome the partial hit they take for changing a term-limits law, but if they are, voters are unlikely to punish them on Election Day.” Elected officials who break a pledge to serve limited terms or vote to repeal term limits, they report, “tend not to face electoral backlash.” In other words, term limits make it difficult to win a third term, but not so difficult that an extraordinarily popular or effective politician can’t pull it off – and that, in the end, may be what voters want.

Will Bloomberg be one of the lucky few who can pull it off? He’s already past the first hurdle, having convinced the City Council to override the referendums of the ’90s and extend the limit to three terms. But he may have to make the sale to the voters, too. Two days before the council vote, Quinnipiac released a new set of polling data. The mayor retained a sky-high 75 percent approval rating. But this time, when asked if they favored extending term limits so they could vote for Bloomberg again, 51 percent of voters said no.

As Maurice Carroll, director of the polling institute, put it: “Who says people have to be consistent?”