Dialing Your Number

The poll and the pollster

by Margaret Tearman

Pollster Bob Carpenter.

It’s dinnertime and the phone rings. You get up, answer it and hear the identifiable pause between your hello and the voice on the other line. You’ve been polled.

Don’t hang up. The caller could well be pollster Bob Carpenter. His living — and a pretty good one it is — depends on your answer. A Washington insider who’s lived in Chesapeake Beach for six years, Carpenter has spent his career on both sides of the polls: first as a poll buyer while working for campaigns and today as vice president of American Viewpoint, a Republican-based polling firm in Alexandria.

Don’t hang up, because more than one man’s income hangs on that dinnertime phone call.

Historic Numbers

Whether you’re asked your voting preference or catching up with the daily news, with Election Day looming, it is nearly impossible to escape the latest political poll. Along with nationally recognized pollsters — like Gallup, Pew and Zogby — it seems every television network, cable news outlet and newspaper is conducting polls.

The partnership between journalism and polling is historic: The first known political poll was conducted by a newspaper. It was 1824 and a presidential contest was underway, with the two frontrunners, Andrew Jackson and John Quincy Adams, battling it out. The Harrisburg Pennsylvanian, a daily newspaper, conducted a straw poll of citizens, by show of hands. The newspaper ran the results, predicting Jackson would beat Adams — and he did, winning the popular vote by 77,000 votes.

But as Al Gore knows, winning the popular vote doesn’t always get you to the White House. In 1824, none of the candidates won a majority of electoral votes and the election was settled in the House of Representatives, where Adams was eventually named president.

In the 1916 presidential election, the Literary Digest, a popular magazine, conducted a first in its series of political polls. Millions of postcards were mailed to subscribers, who correctly predicted Woodrow Wilson’s victory. Literary Digest successfully called the next four presidential elections.

The Digest’s polling methods were based on the theory that the more people polled, or sampled, the more accurate the results. But its methodology was flawed. Its sampling was large but it was, as the pollsters say, self-selected, neglecting a large percentage of voters.

In the 1936 presidential election, that discrimination would be The Digest’s downfall. In that contest, The Digest predicted Alfred Langdon would easily defeat Franklin Roosevelt. But Roosevelt won the election with a whopping 63 percent of the vote.

What happened? The Digest sent questionnaires to subscribers and to people who had auto registrations and phone book listings. In doing so, The Digest failed to include the millions of Roosevelt supporters who didn’t have phones or automobiles — but who voted.

The Literary Digest soon went out of business, but the business of polling was off and running.

Computerizing the Numbers

Today’s polling is a tangle of statistics and mathematical averages.

“Computers have, without a doubt, had the greatest impact on how we conduct surveys,” says Carpenter. “Before computers, everything was done on paper, either by phone or door-to-door or mall intercepts.”

But for all the wonders of technology, phones are still a problem. In the age of the cell phone, many voters — especially young adults who may be first-time voters — aren’t listed in the phone book.

Pollsters are just catching up with cell times. “Today polls sometimes do a sub-sample of cell phone users, primarily young people,” says Carpenter. But your cell phone won’t be called by a computer. “If we’re calling a cell phone, we must use a live operator. State laws prohibit the use of computer-assisted dialing.”

Yes, it is a computer calling home to interrupt your dinner. “If the call is answered, the computer sends the respondent to a live interviewer to complete the interview,” Carpenter says.

Living by the Numbers

Polling was not in the young Carpenter’s numbers. With his undergraduate degree in political science and M.A. in public administration from the University of Southern California, Carpenter built his career on the other side of the polls.

“For 15 years, I did a lot of state party work and political consulting,” he says. “I was heavily involved in several campaigns as a campaign manager.”

In 1994, Carpenter made the move to D.C. and found polling, over a networking dinner with Linda DiVall, president and CEO of American Viewpoint.

Today Carpenter will power-point you through national polls to the nitty-gritty of state calculus in electoral vote totals. He thrives on maps, colored blue, red, pink and purple, the latter two for leaning states.

Carpenter has also jumped the electoral fence, from campaigner to candidate. He’s one of 12 candidates running for the six seats on the Chesapeake Beach Town Council. His campaign has put him back where he started: pounding the pavement, knocking on doors.

“It is different being on the other side, again,” Carpenter says. “I guess the more things change, the more they stay the same.”

Full Circle

It’s Carpenter the pollster who’ll go to work this Election Day.

Once upon a time, pollsters quit a couple of weeks before the election, believing that once voters decided, they wouldn’t change their minds. That mistake was likely a culprit in the infamous 1948 polling fiasco.

For weeks before the election, polls confidently predicted Thomas Dewey would defeat the incumbent, Harry S Truman. The famous photograph of a grinning president-elect Truman holding the Chicago Tribune with the headline Dewey Beats Truman sums up the failure of that mid-century polling.

Current polling continues right up through Election Day, catching voters leaving the polls. These exit polls have come under fire when results have been released before polls have closed. In the 1980 presidential election, NBC predicted a victory for Ronald Reagan at 8:15pm est, based on exit polls of 20,000 voters. It was only 5:15pm on the West Coast, and the polls were still open. There was speculation that voters stayed away after hearing the results, affecting the election. As a result, television networks now hold projected results until after the polls in western states have closed.

“In some cases if a candidate has a significant lead, voters may feel like their vote will not matter so they stay home,” says Carpenter. “But a significant lead may also draw voters to the polls just to prove the pollsters wrong.”

And the Winner Is …

Two weeks from our presidential election, newspapers and television are broadcasting the likely victor — according to the polls. If you believe everything you hear, you may be tempted to skip voting, thinking your vote won’t matter. But history tells us otherwise. Look back 60 years and remember Truman and Dewey.

Public opinion is fickle, and in two weeks, anything can happen. Pollsters know this.

“Pollsters are only as good as the last election,” Carpenter said.

Comments