Why Hillary Clinton could be in even stronger shape than the polls show

Election Day is still a month away, but there are signs that Hillary Clinton might be on stronger footing than even her edge in the polls suggests.

The main reason is that early voting — whether by mail or in person — has already begun in a number of states, and Hillary Clinton’s campaign believes they are winning the early voting battle. On a conference call with journalists Thursday afternoon, Clinton campaign manager Robby Mook said that “states like Nevada, North Carolina, and Florida could be decided before Election Day” based on strong early voting results.

The issue isn’t that a majority of votes will be cast already, but that Clinton can open up an “insurmountable lead” based on high turnout of young and minority voters that, realistically speaking, the Trump campaign cannot overcome.

This could, obviously, just be boasting. But Mook claims, “we are turning out more of our low-propensity voters than the Republicans.” For example, the Clinton campaign says there’s been a 77 percent increase in Florida Hispanics who are requesting a mail ballot and a 79 percent increase in Asian-American requests from that state.

This is playing out in the context of public polling that shows Clinton with a healthy lead and private polling that shows an even healthier one. These indicators suggest the same thing — not only is Clinton ahead, her buffer may be bigger than what current polls show both in terms of public opinion and votes already banked.

A Clinton campaign aide also says that in Iowa so far Democratic returned ballots have exceeded Republican ones three to one, that strongly Democratic Cuyahoga County (Cleveland and environs) accounts for one in six ballot requests received so far in Ohio, and that absentee voting in Northern Virginia has increased at twice the rate of what’s seen in other regions of the state.

The basic backdrop for this is, obviously, Trump’s slide in the polls since his first debate with Hillary Clinton. But a deeper reason for Mook’s confidence comes from the fact that, as Jonathan Martin and Alexander Burns reported for the New York Times, “private polling by both parties shows an even more precipitous drop” for Trump.

Savvy readers know better than to trust selectively leaked private partisan polls over the full suite of information available to the public.

The reality, however, is that campaign polling and media polling methodologies are increasingly diverging, and private polls increasingly have an edge in terms of accuracy. Too few people pick up their phones and talk to pollsters these days to get a statistically valid sample by calling random people. Pollsters make the randomized phone calls, of course, but then they need to do a lot of statistical interpolation to make their phone calls match the overall population.

A modern campaign data operation, by contrast, starts with actual administrative voter lists and then layers upon hundreds of additional layers of commercially available data to build detailed profiles of individuals.

One difference this makes is that if normally Republican-voting members in normally Republican-voting demographic groups — college educated white women, say — become disgruntled with Trump and consequently disinclined to answer phone polls, their place will be filled in a media pollsters’ model with demographically similar people who are still on the Trump Train.

That kind of interpolation could end up masking declines in Trump’s support from key groups.

Alternatively, of course, that kind of thinking could be nothing more than wishful thinking from Clinton supporters. But according to the Times, interviews with “a dozen strategists from both parties” support the view that Trump’s numbers have fallen faster with “independent voters, moderate Republicans and women” than we are yet seeing in the public polls.

Ultimately, nobody will know until the votes are cast.

But Mook’s emphasis on early voting aligns with the reality that Clinton is running a vast campaign based on a cutting-edge data and field operations, while Trump largely is not. Academic research by Ryan Enos and Anthony Fowler indicates that targeted field operations in key swing states boost voter participation by about 7 or 8 percentage points. In a typical election, of course, both sides mount comparable field efforts and the net impact ends up canceling out.

In terms of early voting so far, that may not be the case. Kyle Cheney and Katie Glueck of Politico reported on Tuesday that “Trump’s haphazard campaign, ignoring standard practice, relies largely on mining his boisterous battleground-state rallies to amass his early-vote totals.”

The Clinton campaign, by contrast, has an effort that is “more methodical and traditional, hinging on an extensive field organization to drive its advance voting strategy.”

Mook’s key claim is that so far it is working, with the Clinton campaign succeeding in driving its supporters from lower-turnout demographics to the polls while Republicans are struggling. And what’s clear from all sources — public and private — is that right now Trump is behind in the polls in key states. And with voting already underway Trump has less time to make up lost ground than you might think.