How "The Digest's" Big Poll Was "Put Over"

Literary Digest

THERE WERE SOME 40,000,000 pieces of printed matter in it, a quantity which any statistician could readily figure would cover an appreciable portion of the earth's surface, or, if laid end to end, would reach from here to some other place, and the problem was so to distribute them throughout the nation, get them marked, and get them back, that the tabulated results would give a true reflection of America's feeling toward Prohibition and the soldiers' bonus. How was it done?

The telephone books of the country, which contain the huge total of some 7,800,000 names, and which exhaustive tests have proved to include a representative portion of the population, were chosen as the chief basis of THE DIGEST'S polling list of some 10,000,000 names. Such a list is open to two objections: it contains only 10 per cent. of women to 90 per cent. of men, and it lacks a due proportion of ,workingmen and workingwomen. To even matters up, therefore, THE DIGEST polled half a dozen leading factories of the country and sent out a special women's poll of 2,200,000 ballots. Women voters, in the last national election, were to the men as two-fifths to three-fifths and, counting 10 per cent. of women in the main poll, the supplementary poll virtually balanced the totals as between men and women. These women voters were chosen from voters' lists and city directories. Each State of the Union was given a share, based on its voting strength in the last national election. In most cases, the name of every tenth woman was selected. As in the case of the main poll, it will be seen, the choice was entirely automatic. There was no "hand-picking" either of names or localities.

Photo, captioned workers' votes in sealed boxesThe complete poll, by comparison with the total number of citizens who voted in the last national election in 1920, shows that DIGEST ballots were sent out to more than a third of the total voting electorate. The actual percentage was 37.7. This average was not precisely maintained in all the. States, but inequalities tended to balance each other. Thus, in the Middle Atlantic States, including New York, New Jersey and Pennsylvania, all predominantly "wet," the votes were sent out to 38.7 per cent. of the electorate.

In the much "drier" West South Central States, the percentage turned out to be 48.2. A careful balancing of the whole poll, including "wet"and "dry"sections, justifies the Philadelphia North American's observation that "variations and divergences have a way of neutralizing each other, and experts recognize that in the general average there is an indication of almost uncanny accuracy."

As a safeguard against forgery, the ballots were printed on special machines, which put a layer of color inside each card. By tearing the card, this inner streak of color could be seen. Only a very few machines are prepared to make paper of this sort, and their output was all accounted for during the time

THE DIGEST'S balloting was in progress. Only one attempt at imitation, it appears, was made, and this seems to have been made in good faith. The ballot was a close copy of the regular DIGEST post-card, printed and sent in by a Massachusetts religious society. The total stood 165 for strict enforcement, 11 for modification, and none for repeal. On the "wet" side 80 ballots were caught, all "keyed" alike, and marked by the same railway post-office, at the same hour, apparently with the same pencil, and all voting for repeal. The total of irregular ballots shows 952 for strict enforcement, 248 for modification, and 133 for repeal. The large majority for enforcement is due to the enthusiasm of several "dry." workers who, it appears, copied THE DIGEST ballot more or less perfectly and canvassed for votes among their friends.

Even if all of these irregular ballots had been counted, of course, the result would have been comparable to an extra drop of water in a bucketful.

On a slightly different plane is the vote received on ballots cut from newspaper advertisements and from THE DIGEST'S own pages. In spite of the fact that all of these sample ballots were labeled, in large letters, "Not valid for voting,"a total of 11,963 were received. They stand 4,682 for enforcement, 4,587 for modification, 2,694 for repeal. On the bonus the irregular vote is 6,275 in favor to 5,634 opposed. The opinion disclosed, it will be seen, very closely approximates that of the complete poll. The difference, which may be significant, is in an increase in the vote for enforcement and for repeal, with a de-crease in the vote for modification, thus indicating, possibly, that the people who took the trouble to cut out these ballots, and vote them, felt rather more strongly on the subject than did the general voters.

The same general inference may be drawn from the distribution of votes in the women's poll.

The poll, by and large, has probably aroused more comment and general interest than has ever been aroused by a similar venture undertaken by any American magazine. Both in the number of ballots sent out and returned, it is comparable only to the Presidential Primary undertaken by THE DIGEST two years ago, along similar lines. Papers in England, France and Scotland have carried cables about it, and a subscriber sends us a copy of The Wireless Press News, printed on a steamer en route between America and Japan, containing a radio dispatch bulletined over an area of 46,000,000 square miles of Pacific Ocean, which carries the latest returns on THE LITERARY DIGEST'S Prohibition poll.

A number of correspondents have suggested the use of such a post-card ballot by the Government on questions concerning which information seems to be required of a sort not easily gotten at in the primaries. One suggestion is that such a vote might be taken every three months, the ballot cards to be distributed by the Post-office Department to each voter, to be marked and mailed, franked to Representatives in Washington, who would then be in a position to know what their constituents wish.


No notes

Valid HTML 4.01 Strict Valid CSS2