Session: 1234567891011121314151617         Page 4748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101 of 755

gone down in flames with its polling technique. Elmo Roper, a jewelry salesman who learned something about research by traveling across the country and doing his sales work, started the Roper Public Opinions polling. Knew nothing about statistics. I don't recall how Elmo and I ever met each other but we became very close. And in his 1936 survey of the presidential election that year, he was within less than one percentage point of being right. Gallup was not nearly as close. The Literary Digest was totally wrong. And Elmo and I did a lot of work together on that survey. I didn't do it as a part of his organization -- he had a very small organization. But there were things that I helped him on, and I was certainly close to the tabulation and to the analysis. And we became very close and, in fact, during World War II when OSS was first formed, Bill [William J.] Donovan wanted information not only about information outside the country, he wanted a lot of information about what was going on in public reactions inside the country. It wasn't something that the OSS could provide for him. He didn't know anything about public opinion polling. Roper met him, told him about polling, and he said, “Well, I want polls.” So Roper said to me, “Come on, help me.” And I went to Washington on and off and we helped set up what was then called The Office of Facts and Figures. We were a very small part of setting it up, but at any rate, this was an important wing. And then that operation was moved over to the Office of War Information. And they had a Bureau of Intelligence. And public opinion information was very important about attitudes towards the war and toward rationing and so forth. And the man who headed that, the Office of Intelligence in OWI, was a man by the name of R. Keith Kane. And Kane, who was a Wall Street lawyer, wonderful guy, had a fantastic record at Harvard -- Keith Cane said, “I need help.” And how I met him, I don't know. But I got Gallup and Roper to come advise. And we did polling. And I served as chairman of that three-man committee. And two or three days a week we'd go to Washington and be experts. Everybody was working down there at that time.