Roanoke Times
Copyright (c) 1995, Landmark Communications, Inc.
DATE: TUESDAY, November 8, 1994 TAG: 9411080073
SECTION: EDITORIAL PAGE: A-5 EDITION: METRO
SOURCE: RICHARD MORIN
DATELINE: LENGTH: Long
Who knows? But an article in the latest American Sociological Review answers the question of when horse-race polls began. And it's a fascinating story.
The article recounts the history of a massive ``secret'' poll commissioned by an increasingly anxious Roosevelt brain trust in 1935 to gauge the impact of Louisiana Sen. Huey Long and his Share Our Wealth program on the 1936 presidential elections.
Political scientists Edwin Amenta, Kathleen Dunleavy and Mary Bernstein of New York University note in their article that this survey was the world's first ``scientific public opinion poll on a presidential race,'' coming a year before the more famous effort of George Gallup.
First, a bit of history. Long had established Share Our Wealth in 1934 with the ambitious goal of ending the Depression. He proposed limiting incomes to no more than $3 million through direct taxation. Every American family would be provided a ``homestead'' - about $4,500 - to purchase a ``a house, car and radio - the necessities of middle-class life as redefined by the 1920s,'' Amenta wrote. Special help also was offered to older Americans and World War I veterans.
Needless to say, Share Our Wealth potentially represented a considerable political threat to President Franklin D. Roosevelt, whose first New Deal - those programs passed during his first 100 days in office - had yet to brighten the lives of many Depression-weary Americans. In fact, within a year, there reportedly were 27,431 Share Our Wealth clubs across the country with 7.6 million members.
But just how big a threat was Long? And what, if anything, could the administration do in the next year to blunt his impact? To find out, Democratic strategists decided to do a national poll that was equal parts straw poll and a random-sample survey.
The project was the brainchild of Emil Hurja, chief statistician and an executive director of the Democratic National Committee. Hurja used a two-tiered sampling method to target respondents, drawing his first sample of 100,000 from lists of telephone owners. Since these respondents were generally upscale, employed people, he separately surveyed 49,742 persons who were receiving unemployment benefits. (The sample of unemployed was later ignored by Hurja for unspecified reasons.)
On April 30, 1935, straw ballots printed on a 3-by-5 white cards were sent to the telephone owners sample; the relief recipients were sent blue cards. Overall, 30,924 ballots were returned, a response rate of about 21 percent - significantly better than modern response rates for the first wave of a modern mail surveys. (Pollsters would correctly note that a sample of telephone owners and a 21 percent response rate considerably demolish any claims to randomness, but for its time, this project represented methodological rigor.)
Hurja went to great lengths to hide the poll's real sponsor. The cover letter identified the sponsor as a nonexistent magazine.
The envelope was then mailed to selected potential respondents, most of whom were men. It included a six-paragraph cover letter and the straw ballot. The return postage on the ballot was prepaid, thanks to Roosevelt confidant James A. Farley, chairman of the Democratic National Committee and also the Postmaster General.
The post-card ballot asked the presidential horse-race question in much the same way that it is asked today: ``If an election were held today, who would be your choice of President of the United States,'' and then instructed respondents to mark an X in one of four boxes corresponding to Franklin D. Roosevelt, a Republican candidate, Huey P. Long and a dotted line for a write-in alternative.
When the ballots were counted, their worst fears were confirmed. The results ``indicated a tight race,'' Amenta reported. Roosevelt had 49.3 percent of the vote, the Republican 42.5 percent, and Long with 7.4 percent. (Write-ins accounted for the remainder of the vote.)
``The official results ... show that Long was on the verge of delivering the election to the Republicans,'' the authors wrote. ``If Long was drawing support only from the left, the poll indicated that he would transfer five states, including New York, and 122 electoral voters to the unknown Republican candidate. According to these results, it would take only the defection of Michigan, Iowa, and Minnesota - where Roosevelt held a lead of 7 percentage points or less - to swing the election.''
As a result, Amenta and his colleagues found, the Roosevelt administration began retooling some elements of its legislative agenda for 1935 and 1936, programs that included the bill that became the Social Security Act and were collectively known as the Second New Deal.
Tragedy intervened to prevent a matchup between Long and Roosevelt. In September 1935, Long was assassinated. The Share Our Wealth organization immediately collapsed, and with it Long's dream of ``Share the wealth - every man a king.''
Richard Morin is director of polling for The Washington Post.
The Washington Post
Keywords:
POLITICS
by CNB