Press "Enter" to skip to content

Posts published in “Day: September 19, 2015”

Polls in conflict?


You may have noticed about a week ago reports about two polls of presidential preferences among Idaho Republicans, in separate stories. If you put them together in one story, you can see the results of the two appear to conflict.

But there’s a straight line through them that says something about who supports who.

First, Dan Brown & Associates, from Utah, released a poll of 508 Idaho adults. Among Republicans, businessman Donald Trump took 28% of the vote for the lead. Physician Ben Carson came in second with 15%. Former front-runner Jeb Bush was down in single digits at eight percent; others were in single digits. This was fairly reflective of most of the recent national polls of Republicans (or what you could see in their placement in last week’s presidential debate).

A few days later Republican organizations in Bannock and Jefferson counties tried their own local straw polls, and the results there were a little different. Both counties placed Carson in a strong first place, with about 30% of the vote in each county. In Bannock, Trump was second at 22%, and Florida Senator Marco Rubio came in third at 20%. In Jefferson, Trump was far down the list, as second place went to Senator Ted Cruz (15%), and Senator Rand Paul followed him.

These are distinctly different results, even accounting for the more local polling from the counties. What should we make of these differences?

Here’s some speculation (and if someone from Bannock or Jefferson counties has an alternative explanation, send me a note).

The Brown poll, which was scientifically conducted, probably covered a broad range of Idahoans (other parts of the poll included results among Democratic contenders), and in such a poll party leaders, foot soldiers and activists would account for only a minute portion of the total. It was a “general population” poll.

The straw polls would have been informal, with no specific attempt, as in scientific polls, to account for various percentage portions of the population: The votes they get usually come from whoever happens by. That doesn’t mean these polls are garbage. Years ago as a reporter at the Idaho State Journal I worked with straw polls the newspaper ran at local grocery stores, and when it came to local voting a few days before elections they tended to be surprisingly accurate.

But local people active in the county Republican parties easily could have been over-represented in these two new straw polls.

And that leads to this suggestion:

Among the less-organized, out-in-the-fields Republicans (or Republican-inclined voters) around the state, Trump is highly popular.

Among the more organized Republicans, he may be much less so, with candidates like Carson, Cruz and Rubio finding more appeal. Based on the polls, Bush seems for now to be losing steam in Idaho as he has been nationally.

At least, that looks like a reasonable view from September 2015. Now we can wait a few months and see what it looks like around the holidays.

First take/towns

Ranked lists often pull me in, even when I know better, as here in this list of Idaho's most redneck towns - after all, "redneck" isn't even defined anywhere inn this case. Even if it doesn't have a formal or statistical basis, at least describe your term and what you mean by it. Here are the 10 "most redneck" in order: Salmon, Burley, St. Maries, Ponderay, Jerome, Kamiah, Homedale, Victor, Challis and Weiser. Apart from being in a roughly similar popular span, these places don't really have a lot in common. Some of them are generally remote from urban areas (Salmon, Challis, Kamiah, St. Maries) but others are not far from larger population centers (Ponderay, Jerome, Homedale). Some have a feeling of being under-developed without a lot of new growth or activity, but that sure doesn't fit Ponderday or Victor (just look at the pictures). I know lists like this are mostly just for fun, but come on: If you're going to take the trouble to do something like this, at least define your terms. - rs