Tuesday, August 7, 2012


HUNG UP?: Getting telephone polled (and everyone knows how painful that can be, as Steve Allen used to say) used to be a rare event. While some were annoyed at having to answer questions from a stranger at dinner time, more often "poll-ees" felt that it was so out of the ordinary that they might as well thoughtfully answer the "poll-ers'" questions.

But this year, with the ubiquitous use of the "robo-call"- those pre-recorded "push '1' if you plan to vote for candidate 'X'" calls- we knew it couldn't just be our imagination that the phone seems to ring with these things once a day and twice on Sunday.

We've actually answered at least a half a dozen of theses things this year and screened out another large handful- usually with an "egad, not again" attitude- more than the total number we've ever received over the years.

We'd gotten a strange feeling that these robo-calls- a term that some consider pejorative even though it's the common parlance for any pre-recorded call- were yielding bogus results for a long time, for many reasons.

But the absurdly counter-intuitive results of a "Civil Beat (CB)" poll yesterday that has Tulsi Gabbard taking a sudden and stunning 49%-29% lead over Mufi Hannemann in the 2nd U.S. Congressional District race, all but confirms our suspicions... especially coming, as it does, on the heels of a CB-reported tie a few weeks back and a 10 point lead by Hannemann in a "Hawai`i News Now/Star Advertiser HNN/SA" poll a week before.

Even with the respective "margins of error" a simply statistical explanation doesn't cut it. There has to be something else going on here.

And if our experiences, along with those of some of our Facebook "friends," is any indication it is the robo-calls themselves that provide an explanation.

First of all it appears that only those with a "land-line" have been robo-called. Those with cell phones need not apply. Some say that this discriminates against younger and poorer voters being included but no matter what demographic groups it favors, the results are going to be skewed one way or another.

It may be simply the annoyance factor that causes people to auto-hang-up on auto-calls.

We've developed a habit of doing just that. Whether it's the recent polling or other calls of late, as soon as it becomes apparent that the call is a recording, we hang up,`` having developed the attitude that if it's not important enough for them to call personally it's not important enough for us to answer.

Plus of course there is the modern phenomena of screening calls based on caller ID. This may vary with how busy one is at the moment but if we're busy in the kitchen and an unknown mainland number comes up, we're a whole lot less likely to answer it- and even less so if the caller's number and name are blocked.

We basically asked our Facebook "friends" if they had landlines or cells, how many calls they got and whether the calls were "live or Memorex" ("taped," for all you youngsters) as well as how the number of calls compared to past years.

What we found is that that we're not alone in our response to robo-calls or "touch-tone polling" as Civil Beat would prefer they be referred to. Here are some of the responses left on on our Facebook page (all "Sic"):

- I have been getting at least 3 to 4 a week on land line. I hang up or don't take calls but then they go to message and I still have to deal with them. Most of them come in early evening. They are from everywhere; local, state, fed.

- I think I've probably gotten about 3 in the last month. Not sure. I hang up as soon as I realize its a computer. Been getting a few voice mails asking me to vote for their candidate

-1) hangup on robo-calls, 2) Ask any live person who commissioned the poll, have yet to get an answer and hang-up, 3) if they tell me who commissioned the poll I throw flak at the pollster.

- I have had more survey calls than ever before. Maybe a total of six, four were robo calls. Two were definite push polls, maybe three. All calls were to a land line. Only screened one that I did not take, a repeat call from one surveyor I hung up on for being a push poll. One seemed pointed at prosecutor office, all others combined house and senate. One earlier included county council questions.

In response to that last one, probably because it wasn't clear, candidate for Prosecuting Attorney Justin Kollar responded, "Just so everyone knows, my campaign has done no robocalling. ZERO."

That concurs with a report from Joan Conrow- one that quite a few others have independently confirmed for us- that described what could only be called a "push poll" and seemed to be from Kollar's opponent, incumbent Shaylene Iseri Carvalho.

Our favorite questions from that Iseri push poll was "do you read the blogs?" quite obviously because she's in trouble if they do.

Oh, and here's one response we got via email,

I've had about the same experience you have Andy, though far more than a half dozen, with many calls being from out of the area and automated, a few were actual live interview calls. Some are obviously partisan, others I could not tell who was sponsoring them. I screened one caller, Bob Marx, as noted on my caller ID, and they called back 4 times within a couple of hours. They have all been on my land line, apparently gotten from the phone book. They are a nuisance.

But though this is a small sampling and of course is not a "scientific poll" it is significant in that everyone who responded did so with frustration- oh, all right call it anger- over robo-calls. Remember they don't seem to object to getting polling calls, just that they were the recorded type.

One factor we haven’t heard mentioned is the relative difficulty of the whole "touch tone call" phenomena. Although most of us are used to it by now, no one likes it. Humorists and comedians have had a field day with them for more than a decade now.

Arguably, once you get past the usual desire to talk to a live person- and the fact that that option either doesn't seem to be on the list or is the last one of 10 choices (and is never "0")- the worst part is trying to find the right button on these telephones that no longer have a separate receiver and number pad, but instead have the keypad right next to where you talk and listen so that you have to keep taking the phone away from your head to push the right button and bringing it back to your ear... until your arms (or ear) starts to fall off.

And boy are you in trouble if you didn't catch the first couple of selections of what has become a long, bewildering set of choices. Are you really going to have the machine read the selections again? Or are you more likely to just push any button just to get the thing over with?

And then there's the regular poll that turns into a push poll. Although there's been some argument- usually from the people who run them- that they aren't push polls for one double-talking reason or another, everyone knows them when they hear them.

And they've heard quite a few this year.

Our favorite was one where, five minutes in, the "caller" asks a series of questions beginning with "Would you be more or less likely to vote for Mufi if you knew that he" followed by some horrendous anecdotal tale of apparent avarice or corruption. One question included the words "Pearl Harbor" and "Senator Inouye" and if you weren’t paying close attention you might get the impression that Mufi was at the controls of a Japanese Zero honing in on the 442nd.

Another negative of robo-calls is the impatience factor with a long list of choices. You just know that 10 minutes into a call that had asked for "five minutes of your time" people are starting to push "1" or "2" just to make the process go faster, thinking they already have 10 minutes invested so they don't want to hang up now, they just want it to be over.

Whether these things we do make sense or not isn't important- it's simply what we do without really thinking about it.

Then of course there's the one where the choices change and they type of question remains the same. After a list of four or five questions with the same two possible answers, all of a sudden you realize there are now seven different choices and you just pushed "1" only because that was the one you had been pushing on similar questions.

But the big question here is, assuming that robo-calls give skewed data, why would it favor Tulsi over Mufi?

Well, let's look at those specifics.

Mufi is not exactly the kind of candidate that makes people excited about his candidacy. First of all he's a conservative candidate in what has been arguably described as "the most liberal district in the country." And he's running in a Democratic primary where he started with a big lead without any real progressive oppositions.

The Democratic base had been craving a candidate they could get excited about. Former Senate Majority Leader Gary Hooser apparently chose not to run for the post he ran for some years back and many suspect it was because of the daunting task of raising enough money to take on the presumed favorite, Hannemann. And he had name recognition after also running for lt. governor last year whereas Gabbard has never run for statewide office before, just state representative and Honolulu county council.

Enter Tulsi. Then enter lots of money- much of it from the mainland- to publicize her status as a "born-again lefty."

Even though anywhere else he'd be a Republican, Mufi's support has come from the Democratic "machine" who are supporting him as the "establishment" candidate. Many have been supporting him simply because of name recognition.

But as Gabbard started to raise money and get her name out there she negated much of that and now has become the "go to" candidate who, although they were probably going to vote for Mufi before, has now excited those who were looking for a "progressive alternative."

And make no mistake about it- the Democratic progressives have become excited having long "got over" her former radical right wing stance on marriage rights and other issues.

Those are the people who would be more likely to stay on the phone to get their "choice" registered with the thought that polls numbers create the bandwagon effect.

No one's getting "excited" over the prospect of Congressman Hannemann... except maybe for those who expect a federal contract from the well "know pay-to-play" pol.

And don't forget about the "Mufi haters." Hannemann has tremendous negatives with high "dislike" numbers that he's generated over the years. He's been characterized, if not caricatured, as having a "bullying" style and has been accused of negative- even allegedly racist- campaigning, most recently in his run for governor where he lost to Neil Abercrombie two years ago.

Not too many Mufi Mavens are going to stay on the phone through a robo-call to register their support for him. As a matter of fact, we'd venture to say most of his supporters think it's "in the bag" already- don't forget, this poll wasn't out at the time.

All that could lead to a situation where those most likely to stay through a recorded call- even though they hate robo-calls- would be Gabbard supporters whereas the least likely would be Hannemann supporters.

And for what it's worth, it's a lot easier to hang up on a recording than a person, especially in the "Aloha State" so there's a natural skewing at the most basic level.

We really feel sorry for our friend Mike Levine whose job at Civil Beat was to try to explain how such an anomaly could possibly be valid. He actually did a good job of it and you come away thinking "well, it could happen."

But, upon reflection, no- it couldn't happen. A 30 point difference in two almost simultaneous polls cannot be reasoned away even with the old standby of "it's just snapshot."

When you look at the difference between the HNN/SA and CB polls, the main thing- maybe the only thing- that sticks out is the method of gathering the information. And if it's the method that's in question and one candidate is roundly not just disliked but actively hated then, as most critical thinkers and rational people will tell you, the likely answer is going to be the correct one- that something is kapakahi with the robo-call methodology.

If Tulsi Gabbard beats Mufi Hannemann by 20 points this Saturday we'll eat this keyboard. Because it would be easier to ingest a pound of plastic and metal than it would to swallow the validity of these robo-calls.


Joan Conrow said...

Wonder if Shay's gonna release the results of her poll? ;-)

Blahblahblah said...

54 -- 34

Crunch crunch.

Andy Parx said...

Looks like I'll need yo go out and get 6 gallons of hot sauce today.