A poll released on Thursday, prominently featured at Breitbart News, that shows Donald Trump and Hillary Clinton “trouncing” their rivals “where a victory in the April 19 primary could effectively seal the presidential nomination” was based only on landline calls, is heavily slanted towards voters over the age of 54, and includes independent voters ineligible to vote in the primaries.
Even though the deadline for registered voters to change party affiliation in New York state passed on October 9, 2015, the Emerson poll included independent voters in its findings, and more votes for Trump than the total of eligible Republican voters. New York is a closed primary state, so only voters who are registered as party members are allowed to vote in April, and March 25 is the deadline for unregistered voters to pick a party affiliation and vote at all.
Neither the press release nor the report even mentions if respondents were asked specifically if they were registered to vote.
The poll is unweighted to account for such disparities, and the full report doesn’t include enough breakdowns to determine the ages of voters relied on for each of the primaries or any other claims cited in the press release. None of these facts are mentioned in Washington Political Editor Matthew Boyle’s story that carries the headline in big, bold, block letters — all in caps — at the top of the Breitbart News website.
“Buoyed by a string of decisive primary wins this past Tuesday, both Hillary Clinton and Donald Trump hold a commanding lead in New York State, where a victory in the April 19 primary could effectively seal the presidential nomination for both candidates,” the press release from Boston’s Emerson College Polling Society states.
Boyle’s Breitbart report claims, “A new poll out on Thursday obliterates the latest mainstream media narrative confronting billionaire Donald Trump: That he can’t get majorities, but can only get pluralities, in election results.” Boyle adds, “The delegate rich Empire State offers its 95 delegates proportionally, but if Trump wins this big, he could conceivably get an even bigger slice of the pie than many in the media think he will get.”
Two of the three days included in the poll were before Florida Senator Marco Rubio dropped out of the race Tuesday night, and all of the calls were made on weekdays, when many unretired voters might have been working at their jobs. The press release and accompanying report don’t note the times any of the calls were made, although it claims, “full methodology and results” were provided.
Even some of the Breitbart News commenters were skeptical of the story and poll. “Emerson poll is so full of outliers its hard to make sense of it,” one commenter complained.
“Data was collected using an Interactive Voice Response (IVR) system of landlines only,” according to the full press release, but isn’t noted at the Emerson College website’s extraction, and Boyle doesn’t even mention that in his story.
Sources who have worked at political calling centers say that it’s extraordinarily difficult to reach younger voters on landlines or even cell phones, since they are more apt to screen incoming calls, and disinclined to stay on the line long enough to complete surveys. “I’m in a meeting,” is the top excuse most people who answer the phone before hanging up, when contacted. The report doesn’t even indicate how many calls were made to produce the results. At one NYC calling center, quotas are used, and employees are sent home early if they don’t make enough contacts each hour.
“The GOP primary consisted of 298 likely primary voters, with a margin of error of +/-5.6%, and the Democratic primary consisted of 373 likely primary voters, with a margin of error of +/-5%,” the full press release states.
The numbers noted on the press release are exactly the same as the press release claims, so ineligible voters are included in its findings. The party affiliation listed on the report shows that 373 Democrats, 188 Replicans and 213 Independents were polled.
The Emerson Poll claims that “Trump leads his closest rival, Texas Senator Ted Cruz, by 52 points (64 percent to 12 percent), and that “Ohio Governor John Kasich earned just 1%, getting no bounce from his recent home-field victory in the Buckeye State.” However, there is no breakdown provided in the full report, proving the latter, since it simply provides data from all three days calls were made.
“In the Democratic primary, Clinton is ahead of Vermont Senator Bernie Sanders by 48 points, taking 71% of the vote to Sanders’s 23%,” the press release also claims.
Breakdowns of ages aren’t provided in the press release, but according to the full report only 104 of the respondents were in the 18-34 group and 174 in the 35-54 group. Three hundred and forty are listed as 55-74 and 43 for 75 and older. This accounts for 662 or 85.6% of the respondents, with 111 or 14.4 percent simply listed as “missing”, which may indicate some refused to provide an age or the data was somehow lost. So, according to deducible data, 278 were under 55, while 383 were older, but that isn’t mentioned in the press release.
In an age when younger voters tend to only own cellphones, the report notes that 95 owned only a landline, while 477 of 645 respondents also own a cell phone.
According to the report, 475 intend to vote in the Democratic primary, while 298 will vote in the Republican one, for a total of 773. 263 will vote for Hillary Clinton and 86 for Bernie Sanders, which produced the 71 percent to 23 percent figures claimed in the press release.
The party affiliations noted in the report cited 373 for Democrats, 188 for Republicans, and 213 for Independents.
The raw numbers in the press release show 36 votes for Cruz, 10 for Rubio, 2 for Kasich, and 192 for Trump, which match the numbers noted in the press release.
The press release claims, “Younger voters, ages 18-34, favor Sanders, but by a slimmer margin (53% to 40%) than in many primary states. Among all other age groups, Clinton dominates, leading by 45 points (ages 35-54), 50 points (ages 55-74), and 67 points (ages 75+).” However, that information isn’t provided, at all, in the full report, so it can’t be proven or refuted.