Statistical History of UFOs
Of more than 12,000 sightings eventually registered by the Air Force's Project Blue Book, over 90% were attributed to unremarkable causes such as misidentifications of Venus or weather balloons, or as hoaxes. Surely there are errors in attribution, but activists and skeptics agree that the vast majority of UFO reports indicate nothing extraordinary.
While valueless for physical scientists or engineers, these sightings are useful for sociologists, showing the context in which ET claims occur. For example, UFOs are usually seen after dark but before midnight, and more often in warm months than winter. This reflects the times when people are outside looking at the night sky. Many nations report UFOs, but the United States is the center of activity. Within the U.S. the geographical distribution of sightings correlates roughly with density of non-urban population. Few reports come from urban areas, probably because city lights obscure the night sky.
The Air Force count of UFO sightings ceased with Blue Book’s demise. That loss was remedied by ufologists, one of whom, Larry Hatch, has for 20 years tabulated sightings worldwide and posted them in graphical format on the internet (http://www.larryhatch.net/). Mr. Hatch generously provided me with his raw yearly counts of North American sightings for the years 1947 through 2004. Like Blue Book, Hatch’s unit of analysis is the UFO event, i.e., the sighting of one or more extraordinary objects in the sky, or if on the ground thought capable of flight, at a particular time and place by one or more observers. Sources for his compilation include Blue Book; journals, newsletters, and encyclopedias from UFO organizations; news media; and private catalogs. These pass through his personal filter, weeding out obvious hoaxes, double entries, and misidentified mundane events.[i] I divided Hatch’s yearly count by 811, the maximum count in a single year (1952), producing an indicator of sightings with a maximum value of 1.0.
The New York Times is the nation’s leading newspaper, an agenda setter for other news organs, and the best indexed newspaper during the postwar decades. I tabulated the number of articles about flying saucers/UFOs in the annual New York Times Index from 1947 to 2004 and divided each year’s coverage by the amount in 1966, the year of maximum coverage, producing again an indicator with a maximum value of 1.0. Times coverage correlates highly (r = .78) with yearly counts of magazine articles about UFOs listed in Readers’ Guide to Periodical Literature, adequately indicating years of high and low journalistic attention across the nation.
UFO sightings and Times coverage, both graphed in Figure 9-1, are correlated from year to year (r = .62), rising and falling in concert.[ii] After the burst of saucer sightings and news coverage in early summer of 1947 there was relative quiet in 1948-51, then a sharply defined burst in 1952, and a return to relative quiet in 1953-65. Both indicators peak again -- not always simultaneously -- in 1966-67, 1973, and 1995-97.
[i] Since Hatch’s pre-1970 counts are partly derived from Blue Book counts, it is unsurprising that the two are highly correlated (r = .77) for the period 1948-68. After 1969, with the Blue Book registry gone, the collection of UFO reports may have been less effective, contributing to the subsequent appearance of generally lower counts. This downward counting bias, if it exists, should not affect the detection of post-1969 years with extremely high or low sighting activity.
[ii] Larry Hatch learned of some sightings from news reports, raising the possibility that the indicators are correlated as an artifact of joint measurement. However, since most sightings are not reported in the national news, and most national news stories are not about particular sightings, any methodological conflation must be slight.