home > archive > 2004 > this article

Imperfect polls

By Paul M. Weyrich
web posted October 18, 2004

Public opinion polls burst on the scene in the USA in the 1920's. The popular magazine Literary Digest sponsored the first political polls. The Digest was right on the money in predicting the landslide win of Herbert Hoover in 1928. Likewise, the magazine forecast Hoover's demise in 1932. Then came the 1936 election. Kansas Governor Alf Landon was the Republican nominee. Franklin D. Roosevelt was running for a second term. The 1928 Democrat nominee, Alfred E. Smith, broke with his party and denounced FDR for breaking almost all of his campaign promises in 1932. Literary Digest predicted that Landon would defeat FDR.

Harry Truman holds up a newspaper incorrectly reporting a Thomas Dewey win

Of course, FDR won one of the greatest landslides in U.S. history, all but wiping out the Republican Party. What happened? Literary Digest contacted voters by phone. Those voters who could afford a phone did favor Landon. But the millions of unemployed who were barely hanging on were not polled. Literary Digest was so humiliated by its incorrect prediction that it folded. Public opinion polls in elections didn't surface again until the 1940s, when in 1948 pollsters predicted New York Governor Thomas E. Dewey would defeat President Harry S. Truman. The problem was that Dewey was so far ahead at the end of September that pollsters stopped polling. Thus they completely missed the last minute surge for Truman, largely the result of his "Give ‘em Hell" whistle-stop train tour.

I mention this because political pundit Charlie Cook of the Cook Political Report pointed out that pollsters are missing the 10 percent of Americans who have only cell phones. There is no directory for them to call. It happens that all three people I know who have only a cell phone are Republicans. Not that such a sample means anything, but 10 percent of the population is a large number. What if pollsters, all of whom are predicting a razor-thin race for President, were all wrong because voters owning only cell phones were largely disposed to one of the candidates. That could throw off their calculations big time.

Look at what just happened in Australia. All the pollsters there said the race was too close to call. It wasn't close at all. John Howard, the incumbent, who was supposed to be in deep trouble for deploying troops in Iraq, won an unprecedented fourth term. He picked up a significant number of seats in the lower house and won control of the Senate for the first time during his tenure.

What happened here? Obviously, something significant. Polls were being taken right up to Election Day and none of the pollsters picked up the significant swing to Howard. One possibility is that voters lied to pollsters. The opposition Labor Party had the support of much of the establishment. Perhaps voters didn't want to go against what was perceived to be popular. Some how that seems unlikely. Pollsters often filter out responses they feel are not genuine. More likely, since Australia is so much like America, it is something like the missing cell phone link. I don't have data on how many voters in Australia have cell phones and rely on them exclusively. If it is a significant number that could account for pollsters being so far off.

Perhaps I am living in a fantasy world but some how the Australian situation gives me hope that perhaps the same thing will happen here. The establishment is pro-Kerry and is doing everything it can to elect him. I have never seen anything like this in 46 years of political participation in one way or the other. The tilt in the establishment media is so pro-Kerry it is beyond bias. The media has an agenda and it seems to be willing to do anything to further that agenda. Perhaps some voters are telling pollsters what they think they want to hear. Or it could be the cell phone factor. Wouldn't it be interesting if on election night the race wouldn't be close at all? Wouldn't it be a surprise for all the pundits, ready for a long night, if the election were not razor-thin but the incumbent President were re-elected by a comfortable margin with the race being decided early?

Oh well, I am entitled to my fantasy. Some of the pollsters have become very arrogant. During the primaries they were making predictions that were only off by two-tenths of a percent. Something needs to bring them down to earth. It would be so satisfying if the American people did just that. 

Paul M. Weyrich is Chairman and CEO of the Free Congress Foundation.

Printer friendly version
Printer friendly version
Send a link to this page!
Send a link to this story

Printer friendly version Send a link to this page!

Get weekly updates about new issues of ESR!



1996 - 2005, Enter Stage Right and/or its creators. All rights reserved.