July 16, 2015 — Another new faulty poll, this one from Suffolk University, again posts businessman Donald Trump to a slight lead over the rest of the Republican presidential field. But, like many other surveys that seem to attract sizable media attention, their primary results are not methodologically sound.
The latest Suffolk effort, conducted over the July 9-12 period, interviewed 1,000 adults, again not all registered voters, and produced a Republican primary sample of only 349 respondents. This might be a reasonable sample size for a congressional district, or even a small state like New Hampshire, but certainly not the whole nation.
And, based upon this unreliable data, Suffolk and the media are projecting Trump with what they allude to be a significant lead over Jeb Bush, Gov. Scott Walker, and the other 16 tested GOP candidates.
The percentages give Trump a mere 17-14-8-6 percent lead over Bush, Walker, and Sen. Ted Cruz (R-TX). But, looking at the raw totals, Trump’s 17 percent of the polling sample translates into exactly 60 people saying they would vote for him in their respective state primary or caucus. To again clarify, that’s 60 people throughout the entire nation. Granted, it is more than any other GOP candidate received in this particular poll, but the sample size contains such a larger error factor that it renders the results as being virtually meaningless.
What the poll does verify, however, is that the Republican race is close and there is no clear front-runner among the throng of candidates. These numbers are consistent with other pollsters in that regard despite having such a small sampling universe.
The general election data, however, is more tangible even though the 1,000 respondents include an unknown segment of registered voters. The voter registration question is not asked. Rather, all respondents were treated as if they are voters. Additionally, the operators were instructed to ask for the youngest adult in each contacted residence, which is another unusual feature in the methodology. Of the 1,000 responses, 928 said they were “very likely” to vote in the 2016 election.
That being the case, the Suffolk general election results do merit some attention because the full sample size is at least acceptable for a national poll. Here, Hillary Clinton leads in seven different hypothetical Democratic and Republican pairings. Here are their results:
Hillary Clinton (D) — 46%
Jeb Bush (R) — 42%
Hillary Clinton (D) — 48%
Scott Walker (R) — 37%
Hillary Clinton (D) — 48%
Ben Carson (R) — 36%
Hillary Clinton (D) — 48%
Rand Paul (R) — 38%
Hillary Clinton (D) — 49%
Mike Huckabee (R) — 40%
Hillary Clinton (D) — 46%
Marco Rubio (R) — 40%
Hillary Clinton (D) — 51%
Donald Trump (R) — 34%
As you can see, the results are similar except when Trump is identified as the Republican nominee. In this instance, Clinton fares the best. While she leads all the others, her margins are certainly not insurmountable.
Perhaps the most damaging bit of news to come from the recent Trump media coverage is his admission that he would consider running in the general election as an Independent. This could wreak havoc on the race, and put the GOP nominee in a difficult position.
We remember the 1992 election, when Bill Clinton unseated then-President George H.W. Bush by attracting just 43 percent of the national vote. This was because Independent Ross Perot garnered 19 percent, even though he won no states. Therefore, Clinton was able to score a landslide 370-168 Electoral College victory even though his popular vote total was well under majority status.
Could Trump play such a spoiler role in 2015, and then ironically elect Hillary Clinton? Quite possibly. If Republicans fear anything from this current Trump blip, it should be the quite likely specter of him running as a lone wolf in November.
