Lies, damn lies, and polls

Lies, damn lies, and polls

“There are three kinds of lies: lies, damned lies, and statistics,” Mark Twain was quoted as saying.
Polls can be all three.
Just ask Presidents Alf Landon, Tom Dewey and Hillary Clinton.
Wait, Landon didn’t win in 1936, Dewey did not defeat Truman in 1948 and Clinton, despite being assured she was about to win in a landslide, lost to Donald Trump.
Polls aren’t perfect. They are snapshots in time, and as with all photos, some are sharper than others.
That’s why we in this campaign take the recent poll numbers with concern but not with the assumption that any sample of 500 voters among the nearly 300,000 South Dakotans likely to vote is automatically reliable.
In 1936, President Franklin Roosevelt was seeking a second term, but pollsters said challenger Alf Landon, the governor of Kansas, would sweep him out of office. Literary Digest, one of the leading magazines of the time, guaranteed a Landon win.
Instead, FDR won in a historic landslide. The polls were faulty and Literary Digest was soon out of circulation permanently.
In 1948, President Harry S. Truman trailed challenger Tom Dewey, the former New York gang-busting attorney who was elected governor of his state before losing a close race to FDR in 1944.
The 1948 election was so out of hand, pollsters stopped collecting data and reporters wrote off Truman. On Election Day, the Chicago Tribune didn’t wait for the results before printing a massive Page 1 headline: “Dewey defeats Truman.”
Nobody could convince Harry he was licked. He traveled across the nation, much as Tim has crossed South Dakota, seeking votes and telling voters to send the “do-nothing” Congress a message by electing him.
After winning in a stunning upset, President Truman wore a broad grin as he held up the Tribune with the dead-wrong headline.
You must recall the 2016 election, when the pollsters and pundits said former first lady, New York senator and secretary of state Hillary Clinton would win in a walk.
On Election Night, the networks and major newspapers announced Clinton had a better than 90 percent chance of winning. The voters thought otherwise, electing Donald Trump.

These three presidential elections aren’t the only example of faulty polls. It’s happened in South Dakota, and not that long ago.
In 1994, Bill Janklow was running for a third term as governor after an eight-year respite from public life. He challenged Gov. Walter Dale Miller in the Republican primary.
On June 3, 1994, the Argus Leader, in a Page 1 story, proclaimed Miller in the lead just four days before the election. According to the story, Miller led Janklow 48-41 in an Argus-KSFY poll.
The paper had to eat its words on June 7, when Janklow defeated Miller 54-46. That means the poll was 14 points off.

Tom Lawrence, communications director for the Tim Bjorkman for Congress campaign, worked for the Argus in 1994 and helped cover the gubernatorial race. Lawrence said he was surprised by the poll figures, since Janklow drew large, enthusiastic crowds everywhere he went that year.
The same is true with Tim, who has been met with strong signs of support across South Dakota for the past 16 months. He’s counting on that more than statistics.
Not all polls are wrong, but enough have been to raise questions about their reliability. That’s why we’re moving forward with a head of steam, armed with the knowledge that South Dakota voters will have the final say in Nov. 6.
After all, Landon, Dewey, Clinton and Miller all learned a lesson about polls. Perhaps others this year will, too.