U.S. President Bill Clinton announced last week that he would put off any final decision on national missile defense. Instead, his successor will have to decide whether to go ahead with the controversial program. It is the right decision. The United States should not rush to deploy an antimissile program that threatens to upset U.S. alliances and could trigger another nuclear arms race.

The U.S. government had set a 2005 target date for the deployment of NMD. That date was based on assessments of the ability of would-be enemies to develop their own ballistic-missile technology. To meet the 2005 deadline, the U.S. would have to award contracts to begin by the end of this year construction of the radar that would be the backbone of any missile shield.

In deferring the decision, Mr. Clinton said that he lacked the "absolute confidence" that existing technology can build the NMD system. That is no surprise. To date, the system has failed in two of three attempts to shoot down a missile -- and each time for a different reason. In other words, the Pentagon has been unable to meet its own minimum criterion for declaring the system technologically feasible. And, as critics such as Mr. Theodore Postel of MIT point out, this has occurred despite efforts to rig the flight tests and make "success" as easy as possible. In the real world, that means the system is not feasible -- especially when deployment would cost at least $60 billion. For NMD proponents, however, that does not seem to matter.