Skip to main content
Why Nevada Polls Are Bad

The whiplash induced by the American presidential primary season is harrowing. One week we are inundated by the snowy charms of Iowa and New Hampshire, beholden to king corn and, wrenchingly, dark prince heroin, and the next we are thrown headlong into Nevada, red-rocked, chalk-dry and notched out like an X-Acto knife in the mountains of the West. It is brimming with standing armies of hotel workers, newly arrived immigrants and ill-fated quickie marriages.

What it is not brimming with is polls.

More Politics

In the lead-up to Iowa and New Hampshire’s contests, scores of polls were conducted. In Nevada, where Democrats will caucus on Saturday and Republicans on Tuesday, only five surveys have been published this year. So why, when American political junkies are used to binging on statistics, are there so few Nevada polls?

For starters, when it comes to surveying public opinion, Nevada is still very much the Wild West, and pollsters may be unwilling to gamble their reputations on the state: Nevada is among the hardest places to poll in the nation, with a spotty track record to prove it. Going into the 2008 Republican caucuses, the polling average gave Mitt Romney just a 5-point advantage over John McCain; Romney ended up winning by 38 points. In 2010 when Republican Sharron Angle challenged Harry Reid, then Senate majority leader, for his seat, the polling average showed her beating the incumbent by a 3-point margin; she lost to Reid by nearly 6 points.

According to Mark Mellman, a Democratic pollster who’s done extensive work in the state on behalf of Reid, caucus polling is “excruciatingly difficult” to begin with, but the fact that Nevada’s caucuses are relatively new makes polling them even more fraught with uncertainty. After the 2004 election, the parties moved from primaries to caucuses, and Nevada was bumped up to a higher spot in the primary calendar, a play to incorporate a broader swath of the American electorate in the candidate-winnowing process.

But that means that Nevada populace’s is still pretty unfamiliar with caucusing, making for difficult polling work. “In Iowa, they know what it is, they know what you’re talking about, they know if they haven’t been and if they’re not likely to go,” Mellman said. “In Nevada where it’s a new concept, more and more people are finding out about it each cycle.” In 2008, less than 10 percent of the state’s voting-eligible population came out to caucus (117,000 people turned out on the Democratic side, 44,000 on the Republican) compared with 16 percent in Iowa. Only 1.9 percent of eligible voters — or just under 33,000 people — turned out to vote in the 2012 GOP caucus.

Nevadans are also notoriously hard to pin down for a survey. Many work night shifts, thanks to the state’s thriving hotel and gaming industry, meaning that they aren’t home to answer the phone in the evening, when polling calls are typically placed. That makes a particular difference on the Democratic side of things, where this year’s race shows Hillary Clinton holding a slim lead over Bernie Sanders. “If you’re only polling at night, you’re missing a fair share of electorate that’s works in the casino and entertainment industry that’s highly unionized, that’s very Democratic,” Mellman said.

And given that the state’s population is highly transitory — the bright lights and job prospects on the Strip draw people in like moths to a flame — phone calling can itself be a highly inefficient process. Many people with Nevada cell phone numbers listed in public records, for instance, might no longer live in the state (maybe their burgeoning poker career flopped and they moved home to Oklahoma), while a new resident is less likely to have registered to vote, and thus might not appear on any list from which a pollster might procure relevant information.

So how to improve Nevada’s polling game? Placing calls during both day and night to accommodate the strange hours of the state’s voters is one tactic that Mellman said he’s been using for a decade, and his 2010 polling had Reid winning over Angle when pretty much every other poll conducted in the state showed the opposite. But much of what’s needed might just be a great deal of patience when it comes to Nevada; with each election cycle, knowledge of the caucuses will spread and patterns about who turns out to the contests will likely begin to emerge, helping pollsters model voter turnout.

Time and tide, though, wait for no (wo)man, especially those who want to be president. The excruciating suspense in the West will continue for a while yet.

Clare Malone is a senior political writer for FiveThirtyEight.

Filed under , , , , , ,

Comments Add Comment