Author Archives: Jim Ellis

Florida: Even More Surprises

July 29, 2015 — The Mason-Dixon Polling & Research statisticians surveyed the Florida electorate (July 20-23; 500 likely Florida Republican primary voters; 500 likely Florida Democratic voters) and predictably uncovered some surprising results. Since so many extraordinary political moves continue to unfold in the Sunshine State, the unusual is fast becoming the order of the day.

In the presidential race, results provide an unexpectedly large lead for their former governor, Jeb Bush. The M-D data finds Bush leading the Republican field with 28 percent, followed by Florida Sen. Marco Rubio at 16 percent, Wisconsin Gov. Scott Walker posting 13 percent, and Donald Trump dropping to fourth position with only 11 percent allegiance.

The numbers tell us several things. Jeb Bush, in his home state, enjoys his largest lead and Florida is apparently the only place where he has an advantage that exceeds one or two points. In second place is the state’s junior senator, Marco Rubio, but he lags a dozen points behind.

Continue reading

Iowa, New Hampshire:
Below the Surface

July 28, 2015 — Over the weekend, NBC/Marist College released their recent polls (July 14-21) conducted in the first two presidential caucus/primary states of Iowa and New Hampshire, hence providing some interesting answers to a few new questions.

The pollsters underscore that the sampling period covers the time both before and after Donald Trump made his highly publicized comments about Sen. John McCain’s (R-AZ) war record. Unlike many of the recent public polls that featured extremely small polling national polling samples, the individual respondent universes for these two surveys are acceptable (Iowa: 1,042 residents; 919 registered voters; 342 likely Democratic Caucus attenders, 320 likely Republican Caucus attenders; New Hampshire: 1,037 residents; 910 registered voters; 329 likely New Hampshire Democratic primary voters, 401 likely New Hampshire Republican primary voters).

The glimpse provided between the registered voters and the overall resident sample is also significant. In each state, there is only a negligible difference between how registered and non-registered voter responded.
Continue reading >

The Florida Jumble

July 27, 2015 — The Florida Supreme Court’s order of a partial re-draw of eight congressional districts has turned the Sunshine State’s politics upside down. With one House member already jumping to the Senate race because he will get an unfavorable draw in his Pinellas County district, another representative may be looking to soon follow suit.

Last week, it was reported that freshman Rep. Gwen Graham (D-Tallahassee) is confirming that she, too, is considering entering the open Senate campaign.

In its decision, the state Supreme Court took the highly unusual action of actually stating how the 5th District — that of Jacksonville Rep. Corinne Brown (D) — should be drawn. Currently, the controversial district encompasses a portion of Duval County (Jacksonville), travels southwest to annex part of Gainesville, and then meanders further south to capture African-American precincts in the city of Sanford before moving into Orlando. The configuration has withstood several challenges under the Voting Rights Act over the past two decades, which is why it remains, but the state high court ruled two weeks ago that it did not meet the proper redistricting criteria under the 2010 voter-passed initiative.
Continue reading >

Hillary Hurting

June 24, 2015 — The Quinnipiac University swing state polls attracted a great deal of media attention after their release Wednesday. With Hillary Clinton trailing three different Republicans in a trio of critical swing states, many believe this justifies the sinking feeling many Democrats are experiencing about her electoral chances.

Unlike many of the recent public polls that have captured major media attention, the Q-Poll sample sizes in the three states: 1,231 registered voters in Colorado; 1,236 in Iowa; and 1,209 in Virginia, are strong. The racial demographic segments largely appear sound though the sample is low for Hispanics in both Virginia and Colorado. While Donald Trump has been projected leading national ballot tests in other surveys, Quinnipiac does not include him in their isolated one-on-ones.

Though these polls do appear to have a slight – probably, two to three point – Republican skew, the results continue to reveal some fundamental weakness in Ms. Clinton’s candidacy. These surveys, and others like them, point to two critical areas that consistently cut against her viability as a national candidate.
Continue reading >

Another Awful Poll;
Kasich Announces

July 23, 2015 — The ABC News/Washington Post poll that posts Donald Trump to a 24-13-12 percent lead over Wisconsin Gov. Scott Walker and ex-Florida governor Jeb Bush, et al, joins a long line of bogus polls.

Once again, as we saw with national polls released last week, this latest respondent sample is inherently flawed. The Suffolk University and Monmouth University surveys that we covered five to seven days ago employed miniscule primary voter samples that were far below statistically relevant levels. Therefore, the aggregate polling results became unreliable.

The new ABC/Post poll is equivalently flawed, in similar and different ways. Taken during the July 16-19 period, the pollsters interviewed 1,002 respondents, which is an acceptable number for a general election sample. But, checking their segmentation of political party identification percentages, we again see primary voting samples of between 200-300 for both national parties.
Continue reading >