Experiments undertaken in the days of nuclear testing indicate that underground explosions produce more fallout than explosions on the ground, all other things being equal. The key thing about earth-penetrating mini-nukes, in this regard, is not that they penetrate the earth but rather that they produce relatively small explosions. The smaller the explosion, the less the fallout. If the new and exotic weapon kills fewer people, it is because of its low-yield, not because of its earth-burrowing quality.

Still, even a mini-nuke's fallout would be extensive. An underground explosion of a 3-kiloton nuclear bomb at the Nevada test site in June 1962 produced a huge high and wide cloud of radioactive dust. A December 1970 test of a 10-kiloton bomb detonated in a vertical shaft 900 feet deep produced a fallout cloud that rose 10,000 feet into the air and traveled (albeit in increasingly dissipated form) from Nevada to the Canadian border.

Finally, as Jonathan Medalia notes in a recent report for the Congressional Research Service, the administration's plan for further study of a "Robust Nuclear earth Penetrator" (for which it has requested—and Congress has bestowed—$15 million) does not envision a mini-nuke. John Gordon, then the director of the National Nuclear Security Administration, the nuclear-weapons division of the Energy Department, is quoted as testifying in 2002 that the emphasis of the RNEP program is on "a more standard-yield system. … There is no design work going on low-yield military options."