Gordon Tullock (PS: Political Science and Politics, January 2004, 2) states (about the 1960 presidential election) that:

“Illinois went for Kennedy by a margin smaller than the usual number of votes produced by the Chicago machine from graveyards and empty lots. Texas went to Kennedy by fewer votes than are normally cast by cattle in South Texas.”

This statement may be intended's a joke, but it is unfortunate that there were never any data collected that could either justify or finally set to rest this longstanding myth.

It turns out that there was a brief study done of the reported results of the Kennedy-Nixon contest in Chicago. It did not concern voter registration, but rather the re-canvass of the lever machine results. (Yes, Chicago used lever machine's at the time, so that Tullock's claim that “the ballots were in the possession of the Democratic bosses in Illinois and Texas” is at least half wrong.)

The study, entitled “An Analysis of the Press Coverage of the 1960 Election in Chicago—A Report by Three Distinguished Political Scientists” was done by Herman Finer, J. G. Kerwin, and C. H. Pritchett (a president of APSA) in February, 1961. The study was never formally published, to my knowledge, and I received a copy in mimeographed form from Marie Garber, Maryland's chief elections officer and director of its State Administrative Board of Election Laws. I referenced the study in my chapter on “Computerized Voting” in Advances in Computers, Volume 32, edited by Marshall C. Yovits of Indiana University and published by Academic Press in 1991.

The quote that I sued from the study was the following: “A recheck of the voting machines in Chicago resulted in a net gain of 312 votes for Nixon out of a total of 1,780,000 votes cast—an amazingly accurate reporting of the vote…. The State Electoral Board, composed of four Republicans and one Democrat, certified the Kennedy victory in the state on the ground that there was not sufficient evidence of fraud in Cook County (which includes Chicago to change the canvass.”

It is necessary to point out that, at the time, the governor of Illinois was William G. Stratton, a Republican. While Mr. Stratton would step down in January, 1961, he had been in office since 1953. That gave him plenty of opportunity to fill state-wide offices with like-minded appointees. Clearly, he did so with the State Electoral Board. The ballots from outside of Cook County, where the lever machines were not used, might very well have been in the hands of Republican county bosses, not Democratic ones. As anyone with any familiarity with Illinois knows, downstate is likely to vote Republican.

The contest in Illinois was very close; Kennedy had won by 8,858 votes out of 4.75 million cast. Even if a recount in Illinois would have changed the outcome in that state, Kennedy still would have had enough electoral votes to win the presidency. Kennedy had received 303 electoral votes, Nixon 219, and Harry Byrd of Virginia 15. If Nixon had obtained Illinois' 27, Kennedy would have still had 276, more than the minimum required of 269.

In addition, the contest in Texas was not particularly close. Kennedy had obtained a majority of the votes and an edge of 46,275 out of 2.3 million ballots cast. The plurality was almost exactly 2% of total ballots. The likelihood of a recount overturning the result seems remote. Given the conditions in both states, it would be a reasonable decision for Mr. Nixon not to demand a recount in either. Despite the implication in Professor Tullock's letter, I do not envision Mr. Nixon's biographers claiming that his later behavior was due in any part to the actions of Mayor Richard Daley in 1960. In fact, Mr. Nixon might have been quite pleased with Mr. Daley's actions in the summer of 1968.