Category Archives: Election Analysis

Sanford’s Next Challenge

By Jim Ellis

Sept. 1, 2017 — We all remember former Gov. Mark Sanford’s ignominious exit from the South Carolina political scene in 2010, a year after his international extra-marital affair became worldwide news. His political exile did not last long, as he was able to return to the US House in 2013, winning a special election for the Charleston-anchored congressional district after then-Rep. Tim Scott (R-Charleston) was appointed to the Senate after incumbent Jim DeMint (R) resigned from office.

Considering the way in which Sanford left office, which before engaging in the affair was widely regarded as a successful governorship to the point of him being mentioned as a potential 2012 presidential candidate, his quick return to elective politics was surprising. And, his re-election to the district he previously represented from 1995-2001 proved rather extraordinary.

After winning the 2013 special, Sanford did not even draw a Democratic opponent in the 2014 regular election, capturing 93 percent of the vote against only minor party opposition.

Continue reading

Confirming Polls in
Alabama & Arizona

By Jim Ellis

Aug. 31, 2017 — Two new surveys were released this week that verify trends for two in-cycle Republican US senators, one in a positive manner, the other, negative.

Harper Polling released new data (Aug. 24-26; 800 likely Alabama Senate run-off voters) that basically confirms the last poll we saw in the current Alabama Senate run-off campaign between former state Supreme Court Chief Justice Roy Moore and appointed Sen. Luther Strange: Voter Surveys & Consulting, Judge Moore leading 45-41 percent. According to HP, the former jurist’s lead is now only 47-45 percent.

Last week, two other polls, from JMC Analytics & Polling – a firm that has been polling not only the Alabama Senate race, but also similar campaign situations in Arizona and Nevada during the past week – and Opinion Savvy came to almost identical conclusions but dramatically different from this week’s data: Moore carrying leads of 19 and 18 percentage points, respectively.

The major dissimilarity prevalent in the Harper poll, when compared to any other current survey in the public domain, is their strongly positive favorability index for Sen. Strange. While the Opinion Savvy result found the appointed incumbent languishing in upside down approval territory among Republicans (40:46 percent positive to negative), the Harper data shows him holding a robust 60:24 percent rating, even better than race leader Moore’s 59:26 percent. President Trump scores well among Alabama Republicans in all the released polls, but most particularly Harper’s (87:10 percent).

Continue reading

Nevada: Who Can Tell?

By Jim Ellis

Aug. 30, 2017 — Two new Nevada Republican polls were released Tuesday that differ so greatly it is difficult to confirm which, if either, is accurate.

JMC Analytics and Polling publicized their new Silver State data (Aug. 24-25; 700 likely GOP registered voters responding to an automated survey) that posts challenger and frequent candidate Danny Tarkanian to be running ahead of incumbent GOP Sen. Dean Heller, 39-31 percent, as the two prepare for a competitive 2018 Republican primary battle.

The Heller campaign immediately responded by releasing their Tarrance Group data from earlier in the month (Aug. 14-16; 300 likely Republican primary voters) that finds a completely different result. According to the Tarrance survey, Sen. Heller actually enjoys a comfortable lead over Tarkanian, 55-33 percent.

So, what does this tell us? In looking at both polling methodologies, we can see certain flaws. The JMC poll is automated with the caveat that the sampling group does not necessarily come from the Nevada universe of actual registered Republican voters. Rather, they could be from a larger segment where the respondents to an automated telephone survey are either self-identified Republicans or from geographic areas where GOP candidates normally perform strongly. Notice that the methodology statement language refers to the sample as being comprised of “likely Republican registered voters”, as opposed to the normal “likely Republican (or Democratic) primary voters.”

Continue reading

McConnell Group Counters

By Jim Ellis

Aug. 29, 2017 — Racked by two recent polls that gave former state Supreme Court Chief Justice Roy Moore leads of 18 points or better over appointed Sen. Luther Strange in the Sept. 26 special Republican run-off election, the Senate Leadership Fund, strong supporters of the interim incumbent and closely associated with Majority Leader Mitch McConnell (R-KY), just released their own competing data.

The Voter Surveys & Consulting polling firm (for the Senate Leadership Fund; Aug. 21-23; 601 likely Alabama Republican run-off voters selected randomly from a list of previous GOP primary voters) finds a much closer contest. According to this new data, Judge Moore’s lead is a much tighter 45-41 percent, very different from the 51-32 and 50-32 percent margins that JMC Analytics & Polling and Opinion Savvy, respectively, published during the previous seven days.

Voter Surveys’ conclusion is a net 15 percentage points different than the other post-primary pollsters while surveying virtually the same universe. Dr. Jan Van Lohuizen, a 40-year polling and research industry veteran, conducted the Voter Surveys poll. Despite its wide variance from the previous pair, this latest study may have the most credibility.

Continue reading

Flake Way Down; Data Questionable

By Jim Ellis

Sen. Jeff Flake (R-AZ)

Sen. Jeff Flake (R-AZ)

Aug. 28, 2017 — A new poll was released late last week showing Arizona Sen. Jeff Flake (R) in very poor re-election position, but the polling methodology yields serious flaws. An earlier poll with greater reliability also shows him trending badly, but brandishing upside-down favorability indexes for political subjects is a seemingly routine occurrence for the second pollster.

The Highground Public Affairs consulting firm surveyed the Arizona electorate (Aug. 18-19; 400 registered Arizona voters; 273 self-identified Republicans) and found Sen. Flake to be trailing in both the primary and general elections. Against former state Sen. Kelli Ward (R), who held Sen. John McCain to a 51-40 percent Republican primary win in 2016, Sen. Flake is down by a wide margin, 42-28 percent.

‘Is Flake trailing by a large double-digit margin? It’s hard to argue such solely based upon this poll. Is it reasonable to believe that the senator is behind, however? That answer is yes.’

It is here, however, where the Highground data reveals serious problems. With a statewide sample of only 273 respondents, it is statistically too small to draw any sound reliable conclusions. And, projecting an incumbent Senator with only 28 percent support among poll responders from his own party certainly brings the result into question.

While Flake has also fared poorly in other earlier primary polls, this is the first one where he trails Ward by a significant margin. Because the sample size is a major cause for concern, the larger 14-point support gap between Ward and Flake should be questioned but still must be viewed as at least somewhat relevant. Is Flake trailing by a large double-digit margin? It’s hard to argue such solely based upon this poll. Is it reasonable to believe that the senator is behind, however? That answer is yes.

For the general election, Highground pairs Sen. Flake only with Rep. Kyrsten Sinema (D-Phoenix). After originally saying she would seek re-election to the House, the congresswoman is now admitting that she is seriously considering running for Senate and several Democratic sources believe she is close to announcing her statewide run. According to the Highground poll, Sinema would lead Sen. Flake in a hypothetical head-to-head contest, 40-32 percent. But, an incumbent in any poll barely breaking 30 percent when the statewide polling sample is the size one usually sees for a lone congressional district has to be viewed with a wary eye.

Public Policy Polling completed their Arizona survey in early August (July 31-Aug. 1; 704 registered Arizona voters) and also found Sen. Flake lagging. Though they asked no head-to-head ballot test questions, Flake scored extremely low in his approval rating index. An elected official of either party tallying an upside down favorability rating in a PPP poll is nothing unusual, however. In fact, almost everyone scores in such a manner.

The Flake numbers, however, exceed even some of the more unpopular previously tested elected officials. According to this PPP survey, his approval index is 18:63 percent positive to negative. His problem is exacerbated in that the polling segmentation crosstabs vary little.

Perhaps the senator’s biggest problem is his standing within his own Republican Party. Here, his index is just 22:57 percent, which is of course a horrific intra-party total for an incumbent. Among Independents and Democrats, the numbers are even worse: 17:65 percent, and 15:67 percent, respectively.

There is little difference in how women and men view Sen. Flake. Among women, his favorability is 18:60 percent; men: 18:65 percent. In segmenting by race, the pollsters divided the respondent universe into just three categories: Hispanics (15:49 percent), whites (19:65 percent), and others (21:71 percent).

Though the PPP polls typically skew negative on the approval rating questions, the Flake numbers signal trouble well beyond any likely methodology flaw. Coupled with what can be reasonably drawn from the Highground survey, the clear conclusion is that Sen. Flake has major political problems and is even in danger of losing his re-nomination battle.

This race will continue to attract attention, but it is clear that we are likely headed for both a highly competitive primary and general election campaign in the desert all through next year.