This terrific report comes from long time election activist Marj Creech, who was observing the Democratic Party primary election in South Carolina yesterday. She spent 90 minutes at a precinct that National Exit Pool contracted polling firms Edison/Mitofsky were exit polling. Also note the “curbside voting” provisions made for one voter. Marj is en route to Florida for continued election observation. Join us in wishing her the keenest eye and the best of luck.

I just witnessed how this exit poll, carried out by Edison/Mitofsky, was done in SC. I was in only one precinct in Greenville, SC, and was surprised to find one exit pollster, sitting at a small table (they gave her a child's desk, she said) She simply asked each voter, as they left the voting area if they would like to be part of the "media exit poll." Behind her was a sign saying CNN and other TV stations. She was a very personable black woman, and not pushy at all; she was a little tired and had been there all day from 7 am and would be there till 7 pm, with an hour for lunch. I didn't ask her how much she was paid, but she seemed happy enough to be doing the job. She admitted to me that she did not believe the surveys she was taking would be the same as the results of the election taking place there, so I said, "Well no, but how about if they combine ALL the exit polling places results; don't you think the official results might match then?" She said no because she went to lunch and missed a lot of voters. I don't think she understood what I was saying. She did not know how many polls were being taken across the state.

Approximately one in three voters took a survey form, and sat there and filled it out, by pencil placing a check in the choices. The form was two-sided and had maybe 20 questions. At first she was going to give me a form, but after I left and came back, she decided she had to keep track of all the forms. They were on a pad of maybe 100? forms and she had three pads.  There were too many questions, she told me, and people didn't want to take the time to fill them out. By the way she asked she was biased toward letting people go without filling them out – the size of the seating space and the fact that she was just one person influenced her I think.

The questions on the form (people would occasionally read them to me) were subjective and not just who they voted for, but about their church going/religion and the "Bill factor," simply if Bill's presence in the campaign influenced Hillary's vote. It also asked race, age, and gender. I don't know what else was asked since I never got to read the form. One woman sitting next to me said there were many questions she could not choose any of the choices alone about what influenced her to vote as she did.

She had to CALL in THE RESULTS as soon as the polls closed, probably on her cell phone.

She said they particularly wanted to know the answer to the Bill question. I think she was recording that one separately before putting their forms in the cardboard box. Some folded the surveys, some handed them to her unfolded.

It seemed very clear to me that this poll was to give the TV pundits something to talk about and/or the campaigns to see what went right or wrong and was not intended to give accurate predictions of the race. After the results from various counties started coming in, within an hour after the polls closed, they were able to predict from that.

My precinct was about 66-75% black, judging from the time I was there. I was there only an hour and a half hour, but traffic was definitely picking up from 4:30 to 5:30pm, when I was finally asked to leave, since I had no official capacity to be there. There was an Obama observer also, but I had not filed for observer status. I told the manager (the head poll worker) I was a reporter for a national voters' rights group. He came over when there were no voters and answered my questions.

All the poll workers were also African American, and seemed quite competent in their jobs. Sometimes they also answered my questions. The turn-out was "very high" and "this was a very important election, at least as important as 04." A woman I interviewed in the parking lot said that 04 was a wake-up call, and now people would really come out to vote. As everyone by now knows, Obama won (according to official tallies with 90% in) by 56% to Hillary's 27% and Edward's 17% (OR CLOSE TO THAT). SC was reported to have 30% blacks in the electorate (not specified if that is registered voters or eligible voters).

Another interesting aspect to this polling besides the form format – as opposed to the pollster filling out the form – was that she was ALSO recording demographics on the people who did NOT choose to participate. Here were the categories on these people: *refused/missed *age range *race *gender. So by pooling these data with the date from the surveys, one would have a breakdown on age, gender, and race for all the voters for the day (except for her missed lunch hour and the ones she just missed getting the info down).

In conclusion, it seems that Edison/Mitofsky is doing the kind of exit poll that I thought we the grassroots invented! It's a survey, the pollster asks everybody to participate, those who refuse are noted by demographic, it's only quasi-private, the pollster decides what the questions should be (or in the Mitofsky case, the media) results are tallied informally (by the one pollster in this case). One difference I note is that the results are available QUICKLY, because they are not double-checked or chain-of-custody secured, and they are compiled quickly by someone with a calculator or computer program as they are called in. The cost in the field was for one person for 12 hours per precinct. If she received $10 per hour, that would be $120 for the day. If there were 200 surveys returned (that's low considering the rate of return I saw in one and half hours) and you had 100 pollsters for a state, that's 20,000 surveys and $12,000. (According to the poll's reported results, they had closer to only 2000 surveys which would mean only TEN pollsters for the state! that's only $1200. But I wouldn't think that would be a very good sample for showing fraud.) If you had 100 citizen volunteers for a state, that's 20,000 surveys taken for free. Or 10,000 for 50 volunteers.

How many statisticians to compile 100 phone calls with data? 2 or 3?  Then later they look at the actual surveys to double check and do other number crunching.

But wouldn't it be nice to be able to put out on the internet, "Citizen pollsters find 10 percent discrepancy in reported results in Ohio, (or wherever)," and we report this the same time the TV stations do.  2 hours after the polls close? 1 hour?  Of course we need to plan this and have the forms run off, citizen volunteers lined up, a way to train them, get them the forms and cardboard boxes and simple signs as to who we are. I am getting started now. I hope some others of you see the value of this.

I'll be in FL Tues participating in Mark Adams' exit poll. He is asking for their signatures. We'll see how that works. 

Oh and BTW the machines used were ES&S ivotronics. NO PAPER TRAIL. Also there was NO electronic voter access card given out, just a blue piece of cardboard carried to the poll worker at the machines, after they checked out at the check-in table. They just touch the screen to get started. The manager said they had used these machines for just a few years; he didn't know how long. ALSO something I thought was strange occurred. One poll worked (there were three including the manager) carried a machine, carried horizontally, no legs, out the door to the curb. I looked puzzled so he said, "Curbside voting." A handicapped van was there and a person in a wheelchair was lowered down the side, part way. The poll worker held the machine while this person poked the screen with a stylus of some sort. Now while I applaud this care shown to the differently abled, I wonder at the practice of carrying a voting machine, presumably with other people's votes recorded inside, out of the building. Is the ivotronic the one that has an override button that allows multiple votes to be made? If this were a paper ballot system would the poll worker be allowed to carry the ballot box outside for a voter to place their vote in? Would a provisional ballot have worked for this person? Would it have been harder to mark a ballot by pencil than touch a screen? I guess that would depend upon the person's ability.  NO ONE I talked to stated any concerns about computer voting with no paper trail. Stay tuned for a report about how SC used to hand count at the precincts in the 70's (and maybe later, my eyewitness did it in the 70's). It was all done in one to two hours!