An op­por­tu­nity de­nied

The Covington News - - Opinion - BAR­BARA MOR­GAN

The phone rang the other night dur­ing din­ner. We of­ten have the news on, al­though that’s not good for di­ges­tion. Our satel­lite ser­vice dis­plays on the TV screen the en­tity and phone num­ber that’s call­ing. We saw “Gallup Poll,” and we each bolted for the phone, al­most trip­ping the other. Fi­nally, some­one thought we were im­por­tant enough for a phone call from the world’s old­est polling com­pany. Per­haps our opin­ions would start a pos­i­tive shift in the so­ci­etal fab­ric of this coun­try, even the world. At last, we’d be able to tell a liv­ing, breath­ing hu­man be­ing what we were think­ing, in­stead of sit­ting around talk­ing back to the TV.

I was pray­ing for a chance to voice sup­port for a ban on sell­ing as­sault weapons and thou­sands of rounds of ammo to de­ranged men. I wanted to rail against wrong­way driv­ers, beer bel­lies, ex­posed bra straps and men’s box­ers, cor­po­ra­tions like Char­ter that won’t an­swer phones at cor­po­rate head­quar­ters in St. Louis, flimsy plas­tic shop­ping bags, and politi­cians of both stripes who won’t an­swer a yes-orno ques­tion with­out a dis­course on the ori­gins of the uni­verse. I wanted to ex­press sup­port for boot­ing Todd Akin from the U.S. House of Rep­re­sen­ta­tives Sci­ence and Tech­nol­ogy Com­mit­tee and send­ing him back to Mars from whence he came.

Well, for­get that.

The per­son on the phone line wanted to speak to whomever had re­cently vis­ited a branch of our lo­cal bank on Peachtree Road in At­lanta, and I was the cul­prit. Gallup only wanted to know if the ser­vice was prompt, friendly and ef­fi­cient. That’s all. Dis­ap­pointed is not the word. Re­cently, a friend had a sim­i­lar ex­pe­ri­ence. Gallup was call­ing, and she raced for the phone with hopes like mine. Alas, the call clicked off as soon as she lifted the re­ceiver — Char­ter? — and her hopes were dashed.

You can read on­line as well as I about the his­tory of polling in this coun­try. Back in 1916, a now de­funct pub­li­ca­tion called The Lit­er­ary Di­gest be­gan na­tional sur­veys of read­ers that led to the cor­rect pre­dic­tions of wins by Woodrow Wil­son, War­ren Hard­ing, Calvin Coolidge, Her­bert Hoover and Franklin Roo­sevelt in 1932. By 1936, the sam­pling of read­ers tilted to­ward those more af­flu­ent and Repub­li­can-lean­ing, so the day be­fore the elec­tion, its poll gave the vic­tory to a man named Alf Landon over a Roo­sevelt re-elec­tion. A man named Ge­orge Gallup took a smaller but more sci­en­tif­i­cally bal­anced sur­vey of the pop­u­la­tion and suc­cess­fully pre­dicted Roo­sevelt’s land­slide win. The Di­gest soon died, but Gallup went on to found a world­wide polling in­sti­tu­tion.

The most com­mon polling tech­niques are mailed ques­tion­naires; phone in­ter­views us­ing ran­domly se­lected land­line num­bers; and per­sonal in­ter­views. All have their short­com­ings. Who knows if the per­son to whom a ques­tion­naire is mailed is the one who fills it out? What about re­sponses from all those who don’t bother to par­tic­i­pate? What if it’s about a sub­ject the tar­get isn’t in­ter­ested in? A phone in­ter­view de­pends on some­one an­swer­ing the phone — and many of us don’t these days be­cause of caller ID. It’s il­le­gal in the U.S. to poll cell phone num­bers, so that leaves a lot of po­ten­tial re­spon­dents with crit­i­cal opin­ions but no land lines un­re­ported and un­counted. Then there’s the per­sonal in­ter­view in which re­spon­dents’ an­swers can be skewed by the race, sex or eth­nic­ity of the in­ter­viewer and the lead­ing or­der in which ques­tions are asked. In all three ma­jor types of polling, the con­struc­tion of the ques­tions can influence the out­come, and polls, as we know, are of­ten de­signed to pro­duce a de­sire out­come for the ben­e­fit of who­ever is pay­ing for it.

When polls are re­ported, es­pe­cially dur­ing elec­tion years, some­thing called the mar­gin of er­ror isn’t al­ways in­cluded, but many times, its ef­fect can be sig­nif­i­cant in a close race, like the one we’re watch­ing now. The mar­gin of er­ror is based on the size of the sam­ple. A large sam­ple pro­duces a smaller mar­gin of er­ror, and vice versa. Say the can­di­dates are within two points of each other, but there’s a mar­gin of er­ror of 4 per­cent. That means the leader could be in sec­ond place or it could be an ac­tual tie. In a close race, such dif­fer­ences can make a sub­stan­tial dif­fer­ence on Elec­tion Day.

There was, how­ever, no mar­gin of er­ror when Ge­or­gia House Speaker David Ral­ston read the re­sults of the Demo­crat and Repub­li­can pri­mary poll of vot­ers on the is­sue of a to­tal ban on lob­by­ist gifts to leg­is­la­tors. Three-fourths of vot­ers were un­equiv­o­cal in their sup­port for a ban, so Ral­ston has about-faced him­self and vows to see one passed — or at least stud­ied — in the com­ing leg­isla­tive ses­sion.

Bar­bara Mor­gan is a Cov­ing­ton res­i­dent with a back­ground in news­pa­per jour­nal­ism, state gov­ern­ment and pol­i­tics.

Newspapers in English

Newspapers from USA

© PressReader. All rights reserved.