To paraphrase Mark Twain, reports of God's demise in American science have been greatly exaggerated. In 1916, four in ten scientists said they believed in a personal God and individual immortality. A new survey shows no decline in this figure, after eighty years of science and skepticism. To measure the strength of belief in an era of ascendant science, the eminent researcher James Leuba conducted a landmark survey in 1916. He found a 60 percent rate of disbelief among scientists, and predicted its increase as education spread.

To test that prediction, we replicated Leuba's exact survey to see what scientists believed in 1996. The result: About 40 percent still believe in a personal God and an afterlife. In both surveys, the other 60 percent was divided roughly between 45 percent who rejected the idea of God, and 15 percent who were agnostic.

Leuba's interest in psychology prompted him to insert a secondary question, asking whether respondents who did not believe in personal immortality nevertheless desired it. More than any other, this question probes the tension between intellect and emotion in some scientists. In 1916, 73 percent of the nonbelievers nevertheless desired immortality (1). That hope has dwindled over 80 years, with only about 36 percent desiring today what they otherwise cannot believe in.

Despite the stable 40-60 split in belief-disbelief over 80 years, there has been a significant shift in views held by the three professions surveyed--mathematics, biology, and physics/astronomy.

The 1996 survey showed that scientists in mathematics are most inclined to hold belief in God (44.6 percent). While biologists showed the highest rates of disbelief/doubt in Leuba's day (69.5 percent), that ranking was given to physicists and astronomers this time around (77.9 percent). Higher belief among physicists in Leuba's survey might have been expected in that era, a time when such leading physicists as Lord Kelvin, Robert Millikan, and Sir Arthur Eddington publicly defended religious belief. In our survey we expected the same, especially with some prominent astrophysicists in the 1990s entertaining the anthropic principle (the idea that the universe had to be the way it is in order for humanity to come into being) and certainly Big Bang cosmology--and many prominent biologists sticking hard by Darwinian naturalism. But we were wrong in that expectation. Leuba's 1916 documentation of disbelief had a political impact in his era, particularly his separate data on declining belief among American college students. The populist Democratic politician Williams Jennings Bryan and some conservative Christians seized on the Leuba data in the 1920s to show the social evils of modernism, and its ultimate impact on the nation's morality. They charged that academic scientists were leading college students into disbelief.

While the Leuba data may have caused sensation in 1916 by suggesting that most scientists do not believe in God, the same percentages for 1996 may do the opposite, surprising typical Americans with the indication that as many as 40 percent of scientists are believers in God and an afterlife. Compared to today's computer-driven, mass telephone surveys, Leuba's effort looks quaint. And yet his was among the earliest efforts to apply the science of statistics to sociology. He published his findings in the landmark book, The Belief in God and Immortality, which historian George Marsden recently described as "both an early effort to apply the scientific use of statistics and a pragmatist tract for scientifically created religion."

In two separate mailings, Leuba sent his survey to a total of 1,000 scientists drawn randomly from the 1910 edition of American Men of Science. He received about a 70 percent response. Similarly, we randomly drew 1,000 names from the current edition of the same volume, now called American Men and Women of Science. Our response was about 60 percent. We stuck to Leuba's apportionment: Half biologists and a quarter each in math and physics/astronomy.

On the quirky side, Leuba's survey found that about 20 percent of the scientists who did not believe in God nevertheless believed in personal immortality. In the 1996 response, the breakdown was closer to what might be expected: respondents tended either to believe in both God and immortality, or to reject both. Leuba defined the divine in very conventional terms: "A God to whom one may pray in expectation of receiving an answer." We believed that because such traditional tenets still prevail in American culture, retaining Leuba's 1916 definition of God--hearing prayers and giving immortality--still was the best simple question. If respondents in Leuba's time did not agree with his survey in general--one respondent said, "This is a lot of damned rot!"--we received unsolicited comments that the definition of God did not allow for enough variation. "Why such a narrow definition [of God]?" asked one 1996 respondent, writing in the survey margin. "I believe in God, but I don't believe that one can expect an answer to prayer."