A recent General Accounting Office report on the alleged moving company practice of "weight bumping" - falsifying the weight of goods to increase charges - had come down hard on the Interstate Commerce Commission's contention that the practice is widespread and costs consumers $20 million annually.

GAO found that ICC did not have enough reliable data on which to base the claim that bumping affects 9 percent of all interstate household moves. Last summer ICC conducted investigations in Florida, Arizona and California, checking about 18,000 shipments. The investigation turned up 300 suspected instances of "weight bumps," which worked out to 1.6 percent.

Of that number, only 104 were found to have weights that were inflated by up to 500 pounds. As of March, 36 cases had been closed or recommended for closure for lack of sufficient evidence, 65 were being investigated, and three were being considered for possible criminal prosecution.

When these discrepancies were pointed out to ICC chairman Daniel O'Neal at a Senate subcommittee hearing last week, O'Neal replied that his main purpose in describing weight bumping as widespread was to alert the public to the practice. His view was disputed by the moving industry.

GAO admonished ICC for not using a statistically valid sample in its investigation and recommended that the chairman order a comprehensive study of weight bumping. GOA's other recommendations included unannounced periodic inspections and regulations for weighing stations. The stations should use prenumbered tickets for purposes of control and verification, GAO said.

The Department of Defense, which handles moves of military personnel, was also criticized by GAO. The watchdog agency found that Defense's contention that there was a 13 percent incidence of weight bumping could not be verified because the data had disappeared in DOD computers.

Weight bumping is hard to detect. GAO listed a few ways in which it can be "easily accomplished." These include filling a truck when the fuel tank is empty and weighing it after the tank has been filled, often resulting in a difference of 1,000 pounds.