FWIW, that is precisely the "rule of thumb" that I've always heard.
From: marfra(--nospam--at)opus.co.tt [mailto:marfra(--nospam--at)opus.co.tt]
Sent: Monday, April 03, 2000 9:47 PM
Subject: Vertical Component of Seismiv Loads in Bridges
..One firm with experience in Japan
recommends the use of the acceleration response spectrum for horizontal
motions factored by two thirds. They also recommend directional
combinations for the three possible dominant directions of motion.