The Pittsburgh Press (August 5, 1944)
Background of news –
Midsummer election odds
By Bertram Benedict
A betting commissioner in St. Louis, scene of Governor Dewey’s conference with Republican governors, reports that betting odds on the outcome of the election in November have narrowed, due to “a flood of Dewey money” in recent days, but that the odds still strongly favor the reelection of President Roosevelt. Immediately after the Republican Convention, the odds were 2.5–1 on Mr. Dewey to win; now they are 9–5. Immediately after the Democratic Convention, the odds on President Roosevelt to win were 1–3; now they are 2–5. At the present odds, a $5 bet on Mr. Dewey would win $9; a $5 bet on Mr. Roosevelt would win $2.
Before the days of systematic polls of voters, the betting odds on presidential elections attracted wide public attention and were credited with something of the same “bandwagon” influence that is at present attributed to the polls.
August polls not practical
During the last 60 years, there have been only two really close presidential elections – those of 1888 and 1916 – but the campaign of 1896 promised a close race down to its last six weeks. In each of these years, the midsummer (August) odds favored the loser – Cleveland in 1888, Bryan in 1896, Hughes in 1916. Greater reliance was placed on the September and October odds than on the August odds.
In 1916, shortly after the nominating conventions, the odds quoted by New York betting commissioners favored the election of Hughes by about 2–1. In the final betting in Wall Street, they were 20–17 (Hughes carried New York by 869,115 to 759,426 for Wilson); in other cities, most bets were made at even money.
Similarly, in 1888, the betting was about even just before the voters went to the polls. Cleveland’s chances were supposed to have been injured by the Sackville-West incident, in which the British Ambassador was enticed into saying he thought the Democratic Party was more favorably disposed towards England than the Republicans. In the election, Cleveland won the popular vote (5,540,050 to 5,444,337) but Harrison had a large majority in the Electoral College (233 to 168).
Laws ban election betting
The laws of two states – New York and Florida – provide that persons who place wagers on the outcome of elections shall be barred from voting on Election Day. In New York, betters on elections are classed with vote-buyers and convicted felons; in Florida with duelists. These laws are seldom enforced; officials at polling places seldom have any way of knowing the identity of betters.
When James W. Gerard, former Ambassador to Germany, placed a bet of $20,000 on the reelection of President Roosevelt early in 1936 at odds of 2–1, that fact was widely published in the newspapers. Mr. Gerard attempted to withdraw from the wager when his attention was drawn to Section 152 of the New York election law which bars betters from the polls. Mr. Gerard’s betting commissioner was unable (in August) to obtain odds longer than 6–5 to replace his 2–1 bet, and Gerard paid $3,400 to be released from his commitment. Father Coughlin offered to bet $25,000 to $16,666.66 (odds 3–2) that the votes in Rhode Island of the independent candidate Lemke would exceed the vote of Alf Landon, the Republican candidate, but the wager fell through. The Rhode Island results in November were:
| Roosevelt | 165,238 |
| Landon | 125,031 |
| Lemke | 19,569 |
| Thomas | 929 |